A pavement condition comprehensive evaluation method, system, product and medium

Abstract

A pavement condition comprehensive evaluation method, system, product and medium. The method comprises: collecting real-time pavement scanning data and obtaining pavement patch information, determining the boundary area of the pavement patch and the original pavement, and obtaining the crack data and height difference of the boundary; collecting first and second thermal imaging data at two preset time points, calculating the temperature change rate of the patch and the original pavement, comparing the two to determine the difference value representing the risk of underlying void; generating a surface condition score based on the three-dimensional and image data of the patch, generating a structural integrity score based on the crack, height difference and difference value, and outputting a pavement comprehensive evaluation result in combination with the two scores. The technical scheme provided in the application improves the accuracy of pavement condition evaluation for pavements with potential structural defects.

Description

Technical Field

[0001] This application relates to the field of electronic digital data processing, and in particular to a method, system, product, and medium for comprehensive assessment of road surface conditions. Background Technology

[0002] With the global mining industry undergoing a profound transformation towards automation and intelligence, the construction of smart mines has become a core trend in the industry's development. Conducting condition assessments of critical infrastructure within mining areas, especially mining roads that bear heavy transportation loads, is crucial for ensuring the safe operation of heavy transport vehicles, maintaining the continuity of ore mining and transportation processes, reducing equipment wear and tear, and optimizing road maintenance costs.

[0003] In related technologies, specifically for mine roads Road surface condition assessment typically employs specialized inspection vehicles equipped with multiple sensors. These vehicles integrate high-resolution industrial cameras and 3D laser scanners, enabling them to navigate the complex environments of mining areas and simultaneously acquire high-definition 2D image data and 3D point cloud data of the road surface. By analyzing the image data, surface defects such as cracks and potholes can be identified and quantified. Simultaneously, by processing the 3D point cloud data, key physical indicators such as road surface smoothness and rut depth can be calculated, thereby achieving a comprehensive evaluation of the road surface condition.

[0004] However, to ensure continuous production, mining roads often require rapid, localized emergency repairs, i.e., patching. While these patches restore the road surface to a smooth state in a short time, making it appear intact in images and 3D data, the underlying old defects may not be completely resolved or may re-develop under prolonged traffic loads and temperature effects, leading to structural instability. Therefore, relying solely on surface condition measurements makes it difficult to perceive and determine whether structural problems such as voiding, settlement, or strength reduction have occurred beneath the patches, resulting in low accuracy of the assessment results. Summary of the Invention

[0005] This application provides a method, system, product, and medium for comprehensive pavement condition assessment, which improves the accuracy of pavement condition assessment for pavements with potential structural defects.

[0006] The first aspect of this application provides a comprehensive road surface condition assessment method, the method comprising: The system compares real-time collected pavement scan data, including pavement image data and 3D point cloud data, with pre-stored historical pavement scan data of the same road segment to identify pavement patches; it determines the boundary area between the pavement patch and the original pavement, and obtains crack data based on the image data of the boundary area, and determines the height difference between the pavement patch and the original pavement based on the 3D point cloud data of the boundary area; it collects first thermal imaging data and second thermal imaging data of the pavement patch and the original pavement at a first preset time point and a second preset time point, respectively; it calculates the temperature change rate of the pavement patch and the temperature change rate of the original pavement within a preset distance range around the pavement patch based on the first thermal imaging data and the second thermal imaging data, respectively; it compares the temperature change rate of the patch with the temperature change rate of the original pavement to determine the difference value used to characterize the risk of voidage under the pavement patch; it generates a surface condition score based on the 3D point cloud data and image data of the pavement patch; it generates a structural integrity score based on crack data, height difference, and difference value; and it outputs a comprehensive pavement assessment result by combining the surface condition score and the structural integrity score.

[0007] In the above embodiments, dynamic thermodynamic analysis is introduced to extend the evaluation dimension from the surface to the interior. Specifically, by acquiring thermal imaging data at two different time points and calculating the rate of temperature change, the differences in heat conduction under different structural states are utilized. A pavement patch with a void underneath (i.e., an air insulation layer) cannot effectively conduct heat to the underlying subgrade under sunlight, causing heat to accumulate on the surface. Its heating rate is significantly faster than that of a healthy pavement in good contact with the subgrade. Therefore, by calculating the difference in the rate of temperature change between the patch and the original pavement, the existence of a void beneath the patch can be quantified. This scheme integrates this dynamic thermodynamic result with traditional static geometric measurements (boundary cracks, height differences) to generate a structural integrity score that considers both the surface and the interior. In this way, even if a patch appears intact in images and 3D data, structural defects underneath can be identified, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, before comparing the patch temperature change rate with the original pavement temperature change rate to determine a difference value used to characterize the risk of delamination beneath the pavement patch, the method further includes: From the road surface image data, the average gray values ​​of the road patch area and the original road surface area are extracted respectively, and the gray value difference is calculated. In the preset emissivity correction database that maps the gray value difference of the road surface to the surface emissivity difference, the emissivity deviation correction value corresponding to the gray value difference is queried. The emissivity deviation correction value is used to correct the patch temperature change rate.

[0009] In the above embodiments, the visual color difference between pavement patches and ordinary pavement is quantified by calculating the grayscale difference. Using a pre-set experimental database, the grayscale difference is converted into an emissivity deviation correction value. Finally, this correction value is used to correct the original patch temperature change rate. This correction process subtracts the pseudo-heating effect contributed by darker colors and greater heat absorption from the observed apparent heating rate. This ensures that the corrected temperature change rate more accurately reflects the heat conduction anomalies dominated by structural defects such as underlying voids, avoiding misjudging healthy patches that heat up quickly simply because of their darker color as high-risk defects. This improves the accuracy of pavement condition assessment for pavements with potential structural defects.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, after comparing the patch temperature change rate with the original pavement temperature change rate to determine a difference value for characterizing the risk of delamination beneath the pavement patch, the method further includes: The road surface patch is divided into multiple gridded sub-regions of preset size. Based on the first and second thermal imaging data, the temperature change rate of each gridded sub-region is calculated separately. The temperature change rate of each sub-region is compared with the overall temperature change rate of the road surface patch to identify abnormal sub-regions. The difference between the temperature change rate of each abnormal sub-region and the original road surface temperature change rate is calculated, and the difference is defined as the risk index of the corresponding abnormal sub-region. The spatial coordinates of each abnormal sub-region are associated with the corresponding risk index to generate a dataset containing data points of the abnormal sub-regions.

[0011] In the above embodiments, by meshing the pavement patches and calculating the temperature change rate separately for each independent meshed sub-region, the granularity of the analysis can be reduced. Subsequently, by comparing the change rate of the sub-region with the overall average, sub-regions with abnormal heating rates can be identified, which are also the most likely locations of potential structural defects. Finally, by assigning risk indices containing spatial coordinates to these abnormal sub-regions, the overall risk value is transformed into a two-dimensional risk distribution map that can intuitively display the location, extent, and severity of defects. This allows for the capture of localized structural damage, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, a structural integrity score is generated based on crack data, height difference, and difference values, specifically including: Each data point in the dataset is treated as a graph node in two-dimensional space. Based on the spatial adjacency relationship of the gridded sub-regions, connectivity analysis is performed on all graph nodes to identify one or more independent defect clusters composed of spatially continuous anomalous sub-regions. The average risk index of all anomalous sub-regions within each independent defect cluster is calculated to obtain a cluster risk characteristic value that characterizes the overall hazard level of the independent defect cluster. Based on crack data, height difference, and cluster risk characteristic value, a structural integrity score is generated.

[0013] In the above embodiments, by introducing graph theory analysis, independent defect clusters consisting of spatially continuous anomalies are connected. These independent defect clusters are highly likely to correspond to a structural defect (such as a continuous area of ​​voids). By identifying these clusters and calculating their internal cluster risk characteristic values, a key indicator characterizing the overall risk level of the actual defect area can be obtained. Finally, substituting this more representative risk indicator (cluster risk characteristic value) into the structural integrity scoring model can avoid underestimating local severe defects due to data averaging, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, after identifying one or more independent defect clusters consisting of spatially contiguous anomalous sub-regions, the method further includes: From historical pavement scanning data, extract historical independent defect clusters from the previous assessment period; spatially match each currently identified independent defect cluster with historical independent defect clusters to identify common-origin defects; calculate the geometric centroids of each common-origin defect cluster and historical independent defect cluster to obtain the current geometric centroid and historical geometric centroid; calculate the displacement of the current geometric centroid relative to the historical geometric centroid to obtain the defect expansion vector characterizing the temporal and spatial evolution trend of the corresponding defect; when the direction of the defect expansion vector points to the boundary region of the pavement patch and the magnitude of the defect expansion vector exceeds the preset expansion length threshold, mark the pavement patch as the highest priority risk area.

[0015] In the above embodiments, by identifying defects of the same origin through spatiotemporal matching and calculating the defect propagation vector, the vague concept of disease development is transformed into a quantifiable physical indicator containing direction and velocity. When the propagation vector of a defect clearly points to the weakest structural boundary region where the pavement patch and the original pavement meet, and its propagation speed (the magnitude of the vector) is very fast, it is marked as the highest priority risk based on this development trend. This can identify high-risk damage that may not be the most serious at present, but is most likely to lead to structural failure in the future, thus improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, before calculating the patch temperature change rate of the road surface patch and the original road surface temperature change rate within a preset distance range around the road surface patch based on the first thermal imaging data and the second thermal imaging data, the method further includes: From the 3D point cloud data, the corresponding laser reflection intensity of the patch and the original pavement are extracted respectively. The laser reflection intensity of the patch and the original pavement are compared with the preset dry pavement reflection intensity benchmark value to obtain the wet area and the reflection deviation value corresponding to each wet area. The reflection deviation value is multiplied by the preset temperature compensation coefficient to obtain the temperature correction amount, and the temperature correction amount is added to the corresponding wet area to generate the corrected first thermal imaging data and second thermal imaging data.

[0017] In the above embodiments, by utilizing the three-dimensional point cloud data of lidar, the wet area is located and its degree of wetness (reflection deviation value) is quantified by analyzing the attenuation of laser reflection intensity (wet surfaces absorb light, and the reflection intensity is significantly lower than the baseline value of dry and clean pavement). Then, using a preset temperature compensation coefficient, this degree of wetness is converted into a specific temperature correction value. The heat lost due to water evaporation is calculated and compensated back into the original thermal imaging data in the form of a temperature value. This ensures that the subsequently calculated temperature change rate can accurately reflect the heat conduction characteristics dominated by differences in the internal structure of the pavement, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, the reflection deviation value is multiplied by a preset temperature compensation coefficient to obtain a temperature correction amount, and the temperature correction amount is added to the corresponding humid area to generate corrected first thermal imaging data and second thermal imaging data, specifically including: Within a preset proximity range of each wet area, a reference area with a laser reflection intensity higher than the baseline value of the reflection intensity of a dry, clean road surface is determined. Based on the first thermal imaging data, a first temperature difference between the wet area and the reference area is calculated. Based on the second thermal imaging data, a second temperature difference between the wet area and the reference area is calculated. The difference between the first temperature difference and the second temperature difference is used to obtain a dynamic evaporative cooling correction value. The dynamic evaporative cooling correction value is added to the second thermal imaging data of the wet area to obtain the corrected second thermal imaging data.

[0019] In the above embodiments, by selecting a drier, cleaner reference area next to the wet area as a real-time control group, instead of relying on a preset, fixed compensation coefficient, the actual temperature difference between the wet area and the reference area at two time points (the first temperature difference and the second temperature difference) is calculated to directly measure the true intensity of the evaporative cooling effect at different times. By subtracting these two temperature differences, the additional cooling caused by increased evaporation between the first and second time points can be quantified, i.e., the dynamic evaporative cooling correction value. Compensating this dynamic correction value back into the thermal imaging data of the second measurement is equivalent to offsetting the data distortion caused by changes in the intensity of water evaporation while preserving the original structural thermal characteristics. This avoids misjudging a healthy area that is heating slowly due to increased water evaporation as normal or having an abnormally low temperature, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0020] In a second aspect, embodiments of this application provide a comprehensive road surface condition assessment system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the comprehensive road surface condition assessment system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a road surface condition comprehensive assessment system, cause the road surface condition comprehensive assessment system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a road surface condition comprehensive assessment system, cause the road surface condition comprehensive assessment system to perform the method described in the first aspect and any possible implementation thereof.

[0023] Understandably, the road surface condition comprehensive assessment system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the road surface condition comprehensive assessment method provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. This application extends the assessment dimension from the surface to the interior by introducing dynamic thermodynamic analysis. Specifically, by acquiring thermal imaging data at two different time points and calculating the rate of temperature change, it utilizes the differences in heat conduction under different structural states. A pavement patch with a void underneath (i.e., an air insulation layer) cannot effectively conduct heat to the underlying subgrade under sunlight, causing heat to accumulate on the surface. Its heating rate is significantly faster than that of a healthy pavement in good contact with the subgrade. Therefore, by calculating the difference in temperature change rates between the patch and the original pavement, the existence of a void beneath the patch can be quantified. This scheme integrates this dynamic thermodynamic result with traditional static geometric measurements (boundary cracks, height differences) to generate a structural integrity score that considers both the surface and the interior. Thus, even if a patch appears intact in images and 3D data, structural defects beneath it can be identified, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0025] 2. This application quantifies the visual color difference between pavement patches and ordinary pavement by calculating grayscale differences. Using a pre-defined experimental database, the grayscale difference is converted into an emissivity deviation correction value. Finally, this correction value is used to correct the original patch temperature change rate. This correction process subtracts the pseudo-heating effect contributed by darker colors and greater heat absorption from the observed apparent heating rate. This ensures that the corrected temperature change rate more accurately reflects the heat conduction anomalies dominated by structural defects such as underlying voids, avoiding misjudging healthy patches that heat up quickly simply because of their darker color as high-risk defects. This improves the accuracy of pavement condition assessment for pavements with potential structural defects.

[0026] 3. This application achieves a more granular analysis by meshing the pavement patches and calculating the temperature change rate for each independent meshed sub-region. Subsequently, by comparing the change rate of each sub-region with the overall average, sub-regions with abnormal heating rates can be identified, which are also the most likely locations of potential structural defects. Finally, by assigning risk indices containing spatial coordinates to these abnormal sub-regions, the overall risk value is transformed into a two-dimensional risk distribution map that visually displays the location, extent, and severity of defects. This allows for the detection of localized structural damage, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating a comprehensive road surface condition assessment method in an embodiment of this application; Figure 2 This is another flowchart illustrating the comprehensive road surface condition assessment method in the embodiments of this application; Figure 3This is an exemplary hardware structure diagram of the road surface condition comprehensive evaluation system in this application embodiment. Detailed Implementation

[0028] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to and includes any or all possible combinations of one or more of the listed items.

[0029] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0030] In related technologies, the condition assessment of mine roads mainly relies on the measurement of the physical surface of the road surface. For example, inspection vehicles equipped with industrial cameras and 3D laser scanners collect images and 3D point cloud data to identify surface cracks, potholes, or calculate indicators such as smoothness. However, this assessment method can only measure the physical surface of the road surface. For road sections in mining areas that have undergone rapid emergency repairs to ensure production continuity, the newly laid patches often appear intact in images and 3D data, but this precisely masks the true structural health beneath. Existing defects may not have been eradicated and, under long-term traffic loads and temperature effects, may develop into hidden structural defects such as voids and settlement, which are undetectable by surface sensors. Therefore, assessing such road sections solely based on surface measurements will result in reduced accuracy because it cannot reflect the true structural stability.

[0031] In this embodiment, to address the problem that surface measurements alone cannot detect internal defects, a dynamic thermodynamic analysis method is introduced, extending the evaluation dimension from the surface to the interior. The core of this method lies in acquiring thermal imaging data at two preset time points and calculating the difference in temperature change rate between the pavement patch and the surrounding original pavement. The physical principle is that patches with underlying structural defects such as voids, due to impaired heat conduction with the roadbed, will heat up significantly faster under sunlight than structurally intact pavement. The calculated difference becomes a key indicator that allows for insight into the condition beneath the patch and quantifies the risk of voids. Finally, this application integrates this dynamic thermodynamic analysis result with traditional static geometric data such as boundary cracks and height differences to generate a structural integrity score. This allows for the identification of potential structural defects even when the patch surface appears intact, improving the accuracy of the evaluation results.

[0032] Figure 1 This is a flowchart illustrating the comprehensive road surface condition assessment method used in the embodiments of this application, including the following steps: S101. Real-time acquisition of road surface image data and 3D point cloud data, and obtaining the operation records of road repair vehicles through V2V broadcast within the work area to obtain road surface patches.

[0033] Among them, road surface scanning data refers to road surface condition information acquired through sensing devices such as laser scanners and cameras, including image data and 3D point cloud data; image data represents the two-dimensional visible light information of the road surface, used to identify surface defects such as cracks and potholes; 3D point cloud data refers to the three-dimensional morphological information of the road surface recorded in spatial coordinates, used to characterize the elevation changes and geometric features of the road surface; the intranet of the work area refers to the local communication network within the road construction area, used to realize information interaction between construction vehicles; V2V broadcast refers to the information broadcast communication method between vehicles (Vehicle to Vehicle); and the work record refers to the information such as construction location, time, and scope recorded by the repair vehicle during the construction process.

[0034] Specifically, by using sensors installed on the inspection vehicle to collect road surface scanning data in real time, image data and 3D point cloud data containing road surface features are obtained. At the same time, the inspection vehicle receives V2V broadcast information from the repair engineering vehicle through the intranet of the work area to obtain the construction operation record. Based on the location information in the operation record, the specific location and range of the road surface patch are located and confirmed in the collected road surface scanning data.

[0035] In some embodiments, road surface data acquisition and patch identification can be achieved in multiple ways: Optionally, data acquisition can be performed using a vehicle-mounted mobile measurement system (MMS): a high-resolution linear or area array industrial camera is installed to acquire road surface texture information; a lidar scanner is configured to acquire the three-dimensional geometric shape of the road surface; a combined navigation system of a global navigation satellite system and an inertial measurement unit is used to assign time-stamped geospatial coordinates and attitude information to each frame of image and point cloud data; a complete road surface scanning dataset is formed through data preprocessing and integration; Optionally, patches can be identified by comparing real-time acquired data with historical data: first, coarse registration is performed using the geographic coordinates provided by the navigation system to align the two acquired data to the same coordinate system; then, fine registration is performed using an iterative nearest-point algorithm to optimize the translation and rotation matrices of the real-time point cloud; based on the aligned data, the height difference of the point cloud is calculated to generate a height difference map, and the appearance texture changes are analyzed to generate a saliency map; finally, the change detection results of geometric and appearance dimensions are fused to determine the patch boundary and output vector data. It is understood that other combinations of data acquisition devices or other data processing and comparative analysis methods can also be used to identify road surface patches, which is not limited here.

[0036] S102. Determine the boundary area between the road patch and the original road surface, obtain crack data based on the image data of the boundary area, and determine the height difference between the road patch and the original road surface based on the three-dimensional point cloud data of the boundary area.

[0037] Specifically, taking the vectorized polygon data that identifies the location of the pavement patch from the previous step, features are extracted from the pavement patch boundary. A pre-defined distance (e.g., 10 cm) is then extended outwards from the center of this polygon boundary line to both the inner and outer sides, thereby generating a ring-shaped boundary analysis zone.

[0038] The image portion covered by the aforementioned boundary analysis zone is cropped from the road surface image data. Image segmentation algorithms are then used to detect cracks in this region, typically employing a deep learning-based semantic segmentation network (such as the U-Net architecture). This network, trained on a large number of road surface crack samples, can identify crack portions in the image at the pixel level and generate a binary crack mask image. Finally, skeletonization and geometric analysis are performed on this crack mask to quantify a series of crack data, such as the total crack length in the boundary region, the average crack width, and the crack density (crack length per unit area).

[0039] Using the acquired 3D point cloud data, the point cloud data within the boundary analysis zone was divided into two subsets: the inner point cloud located inside the patch polygon and the outer point cloud located outside. The average height value of all points in the inner point cloud and the average height value of all points in the outer point cloud were calculated separately. Finally, the height difference between the pavement patch and the original pavement was defined as the difference between these two average height values.

[0040] S103. At the first preset time point and the second preset time point, collect the first thermal imaging data and the second thermal imaging data of the road patch and the original road surface, respectively.

[0041] Specifically, the data acquisition operation is performed in two steps. At the first preset time point, a vehicle is detected driving at a preset speed across a road segment containing the target pavement patch. During this process, the thermal imager continuously records the thermal infrared radiation of the road surface and generates a thermal image with geographic coordinates by fusing it with data from the Global Navigation Satellite System / Inertial Measurement Unit. Using the patch boundary coordinates determined in the previous step, thermal images belonging to the patch and its surrounding area are cropped and saved from the data stream, and this is recorded as the first thermal imaging data. At the second preset time point, the above acquisition process is repeated using the exact same equipment and path, and the acquired data is recorded as the second thermal imaging data.

[0042] The first preset time point is set as the starting point of the road surface thermal state change cycle, which can be set to a very short time after the road surface patch is detected; the second preset time point (T2) is set as the ending point of the change cycle, that is, after the road surface has undergone sufficient heat exchange, which can be set to the moment when the vehicle is about to leave the area corresponding to the road surface patch.

[0043] S104. Based on the first thermal imaging data and the second thermal imaging data, calculate the patch temperature change rate of the road surface patch and the original road surface temperature change rate within a preset distance range around the road surface patch, respectively.

[0044] Specifically, based on the vector boundary of the patch, the patch analysis area is delineated in both sets of thermal imaging data. At the same time, using the patch boundary as a reference, a preset distance (e.g., 1 meter) is extended outward, excluding the patch itself, to generate a ring-shaped original road surface reference area.

[0045] For the patch analysis area, all pixel values ​​in the first thermal imaging data are extracted and their average temperature is calculated; similarly, the average temperature in the second thermal imaging data is calculated. The patch temperature change rate is calculated by subtracting the values ​​from the first preset time point and the time difference between the second preset time point and the first preset time point.

[0046] The exact same calculation process was applied to the original pavement reference area to obtain the original pavement temperature change rate.

[0047] In some embodiments, the road surface may be locally wet or have standing water (e.g., after rain or after watering the roadside). To eliminate the interference of the cooling effect of water evaporation on the thermal imaging data, a data correction process based on the linkage between laser reflection intensity and thermal data can be performed first to ensure the accuracy of subsequent structural assessment.

[0048] Laser reflection intensity is used to accurately identify wet areas. Asphalt and aggregates that make up road surfaces have relatively high reflectivity in the near-infrared band commonly used by lidar, while water is a strong absorber in this band. Therefore, the laser reflection intensity value is first extracted from each data point in the 3D point cloud data for both the road patch and the original road surface area. Simultaneously, a baseline value for the reflection intensity of a dry, clean road surface, calibrated through extensive experiments, is built-in. The actual reflection intensity of each point is compared with this baseline value. When the reflection intensity of an area is significantly lower than the baseline value, and the deviation exceeds a preset threshold, that area is marked as a wet area.

[0049] A localized dynamic thermodynamic reference system is established. For each identified wet area, a reference area with a laser reflection intensity significantly higher than the dry road surface baseline is searched within its immediate vicinity. Selecting an area with an intensity higher than the baseline ensures that the reference point is dry, thus forming a control group. Choosing an adjacent reference area ensures that the wet area and its reference area receive the exact same solar radiation and experience the same wind speeds and other environmental conditions on a macroscopic scale.

[0050] The dynamic evaporative cooling effect is quantified and separated. Calculations are performed using thermal imaging data from two time points. First, in the first thermal imaging data, the temperature difference between the humidified area and a reference area is calculated and denoted as the first temperature difference. This temperature difference primarily reflects the cooling effect caused by water evaporation in the initial state. Then, in the second thermal imaging data, the temperature difference is calculated again for the two identical areas and denoted as the second temperature difference. By subtracting these two temperature differences again, the dynamic evaporative cooling correction value is obtained: second temperature difference - Δfirst temperature difference. This correction value represents the additional temperature drop caused by the dynamic process of evaporation during the time period from the first preset time point to the second preset time point, thus separating the interference of the non-structural factor of water evaporation on the rate of temperature change in a specific numerical form.

[0051] The thermal imaging data is then corrected. Finally, the dynamic evaporative cooling correction value obtained in the previous step is added back to the second thermal imaging data corresponding to the humid area. The physical meaning of this step is equivalent to "drying" the road surface at the data level. After correction, the temperature data of the area that was originally slow to heat up due to moisture is restored to a level that can truly reflect the thermal performance of its materials and the underlying structure.

[0052] The above steps identify wet areas of the road surface using laser reflection intensity data and, combined with thermal imaging data from two different time points, establish a dynamic temperature difference reference between the wet area and the adjacent dry area. By calculating the change in temperature difference, the abnormal cooling effect caused by accelerated moisture evaporation is quantified and separated. Finally, by compensating for the temperature drop caused by this effect back into the original data, the interference factor of moisture can be eliminated without affecting the actual structural thermal characteristics. This improves the accuracy and reliability of subsequent temperature change rate calculations, further enhancing the accuracy of pavement condition assessment for pavements with potential structural defects.

[0053] S105. Compare the patch temperature change rate with the original pavement temperature change rate to determine the difference value used to characterize the risk of voiding under the pavement patch.

[0054] Specifically, the void beneath the road surface is filled with air, which is a poor conductor of heat (i.e., an insulator). Therefore, during daytime warming, a patch with a void underneath will receive less heat from the bottom and its downward heat loss will be hindered. However, this mainly manifests as a slower overall warming rate compared to a healthy road surface that is structurally dense and tightly bonded to the base layer.

[0055] Based on the above principles, the difference value can be calculated using the direct difference method, defined as the arithmetic difference between two rates of change: Difference value = Original pavement temperature change rate - Patch temperature change rate. Under this definition, a positive difference value indicates that the patch's heating rate is slower than the surrounding healthy pavement; the larger the difference value, the more severe the thermal hysteresis, and the higher the risk of voiding beneath the patch. A value close to zero or negative indicates good structural contact beneath the patch.

[0056] In some embodiments, the difference value can also be calculated using the following first formula: Where D is the difference value, and R is... O R represents the rate of change of the original road surface temperature. P Let α be the patch temperature change rate, ΔH be the height difference, W be the average width of cracks detected in the patch boundary region, L be the total length of cracks detected in the patch boundary region, and α and β be dimensionless empirical weighting coefficients. These two coefficients need to be calibrated using a large amount of experimental data to determine the influence weight of height difference and boundary cracks on the final risk assessment. For example, by detecting patches known to have different degrees of voids, the optimal values ​​of α and β can be backfitted.

[0057] By employing the first formula described above, risk assessment is elevated from a single-dimensional to a comprehensive, multi-dimensional judgment. The core part of the formula... The thermal hysteresis rate of the patch relative to the original pavement is a normalized indicator. It eliminates the difference in the absolute value of the temperature rise rate caused by different weather conditions (such as total solar radiation intensity) on different testing days, making the calculated fundamental thermodynamic differences comparable under any conditions. The multiplier term in the formula... Height difference is introduced as a risk amplification factor. If there are structural problems beneath a patch (such as slight settlement due to voids or extrusion bulges caused by base layer defects), it is highly likely to be reflected in the three-dimensional morphology. The larger |ΔH| is, the more severe the deformation in the physical structure. Using this physical deformation as a multiplier means that a patch with both thermal anomalies and significant height difference deformation is far more risky than a patch with only thermal anomalies but a smooth surface. This makes the assessment results more consistent with engineering reality. Another multiplier term in the formula, 1+β×L×W, introduces boundary cracks as another risk amplification factor. Boundary cracks are the main channel for external moisture intrusion beneath the patch, and moisture intrusion is the main cause of base layer erosion, void formation, or weak interlayers. Therefore, the severity of boundary cracks (characterized by the product of length and width) is related to the probability of future void occurrence and development. Incorporating this causal indicator into the model allows the difference value to not only assess the current situation but also predict future deterioration trends to some extent.

[0058] S106. Based on the 3D point cloud data and image data of the road surface patch, generate a surface condition score.

[0059] Specifically, multiple sub-indicators are extracted from both three-dimensional and image dimensions, and then fused into a final surface condition score using a weighted average method.

[0060] Smoothness / Roughness: First, an optimal reference plane is fitted within the 3D point cloud of the patch area. Then, the perpendicular distance (residual) from each point cloud point to this reference plane is calculated. The standard deviation or root mean square error of all these distances is used as a metric for the macroscopic smoothness of the patch surface.

[0061] Rut depth: A virtual ruler is simulated across the cross-section of the patch. The rut depth is determined by calculating the maximum vertical distance between the point cloud surface contour line and this virtual ruler. This calculation is performed multiple times along the driving direction within the patch area, and the average or maximum value is taken as the rut index.

[0062] Texture Wear: New asphalt pavements possess rich macro-texture (gaps between aggregates), which is crucial for skid resistance. Image processing techniques, such as Gray-Level Co-occurrence Matrix (GLCM) analysis or Fourier transform, are used to quantify the texture complexity of patch images. Over time, as vehicles wear down the aggregates, the texture becomes blurred and smoothed. The degree of texture wear is calculated by comparing the current texture feature values ​​with a baseline value representing the texture features of an ideal new pavement.

[0063] Surface defects: Using a specially trained deep learning object detection or image segmentation model (such as YOLO or SegNet), surface-specific defects such as looseness, oil bleeding, and cracking are automatically identified and quantified in the patch image data, and the percentage of each defect area relative to the total patch area is output.

[0064] Finally, all the aforementioned quantitative indicators (smoothness, rut depth, texture wear, and the percentage of various types of defects) are integrated using a pre-defined scoring model. This mathematical model is typically a weighted summation formula: Surface condition score = w1*f1 (smoothness) + w2*f2 (rut depth) + w3*f3 (texture wear) + w4*f4 (percentage of various defects) + ... Here, w is the weight of each indicator, and f is the conversion function that maps the original physical quantities (e.g., meters, millimeters) of each indicator to a standard score range. These weights and conversion functions are defined by road engineering experts based on the severity of the impact of different defects on pavement performance.

[0065] S107. Generate a structural integrity score based on crack data, height difference, and difference value.

[0066] Specifically, a non-linear fusion model based on a deduction system is adopted, where the score starts from an ideal full score (e.g., 100 points) and then the corresponding penalty points are subtracted according to the severity of each defect.

[0067] Structural integrity score = 100 - boundary defect penalty + geometric deformation penalty + subsurface defect penalty.

[0068] The boundary defect penalty term quantifies the risk arising from unclosed boundaries. Boundary cracks are direct channels for moisture penetration and the starting point for structural failure. Boundary defect penalty term = w5(1-e^(-k×L×W)) Where w5 is the weight and k is the adjustment coefficient. This formula shows that once a crack appears, the risk (deduction) increases rapidly, but when the crack is already very serious, its marginal impact tends to level off. It accurately simulates the qualitative change from a closed to an open state.

[0069] The geometric deformation penalty term quantifies the physical deformation caused by structural instability (settlement or uplift). Geometric deformation penalty term = w6(|ΔH| / H1)^γ.

[0070] Where w6 is the weight, H1 is the maximum height difference threshold allowed by engineering specifications, and γ (usually greater than 1) is the exponential coefficient. This formula means that when the height difference |ΔH| is much smaller than the threshold, the penalty is small; but once it approaches or exceeds the threshold, the penalty will increase exponentially to reflect its serious threat to driving safety and comfort.

[0071] The subsurface defect penalty term quantifies the most dangerous underlying structural defects that are invisible to the naked eye. Subsurface defect penalty term = w7(e^(λ×D)-1) w7 is the weight, and λ is the risk sensitivity coefficient. This formula utilizes the previously calculated comprehensive difference value D. The exponential function e^(λ×D) ensures that as the value of D (risk of detachment) increases linearly, the probability of structural failure and the penalty score it represents will increase explosively, which is consistent with the evolution of structural defects from the incubation period to the rapid failure period.

[0072] By adding these three separate penalty items together and subtracting them from the full score, a structural integrity score that comprehensively reflects the health of the structure is obtained.

[0073] S108. Combining surface condition score and structural integrity score, output the comprehensive pavement evaluation result.

[0074] Specifically, a maintenance decision matrix, also known as the four-quadrant analysis method, is used. This method constructs a two-dimensional decision space by using surface condition scores and structural integrity scores as two orthogonal coordinate axes. Typically, the structural integrity score is used as the X-axis, and the surface condition score as the Y-axis. Each axis is further divided into "high" and "low" intervals based on a threshold (e.g., 80 points), thus forming four quadrants. The coordinates of a specific patch (surface condition score, structural integrity score) are plotted into this matrix; the quadrant it falls into directly determines the final comprehensive assessment result and maintenance strategy.

[0075] First quadrant (high surface condition score, high structural integrity score), assessment result: patch surface is intact, structure is stable.

[0076] Quadrant 2 (low surface condition score, high structural integrity score): The surface appears intact, but there is a serious risk of voids or structural separation underneath. It is recommended to use ground-penetrating radar or core sampling for accurate verification and to plan structural repair solutions such as excavation, refilling and compaction.

[0077] The third quadrant (low surface condition score, low structural integrity score) assessment result: the patch surface is damaged and the underlying structure has also failed. This patch is beyond repair and should be completely removed and repaved as soon as possible.

[0078] Quadrant 4 (High Surface Condition Score, Low Structural Integrity Score): The patch structure is solid, but the surface material is worn, cracked, or has other defects. It can be included in the routine maintenance plan and low-cost surface treatment measures such as milling overlay, crack grouting, and micro-surfacing can be taken.

[0079] In the above embodiments, dynamic thermodynamic analysis is introduced to extend the evaluation dimension from the surface to the interior. Specifically, by acquiring thermal imaging data at two different time points and calculating the rate of temperature change, the differences in heat conduction under different structural states are utilized. A pavement patch with a void underneath (i.e., an air insulation layer) cannot effectively conduct heat to the underlying subgrade under sunlight, causing heat to accumulate on the surface. Its heating rate is significantly faster than that of a healthy pavement in good contact with the subgrade. Therefore, by calculating the difference in the rate of temperature change between the patch and the original pavement, the existence of a void beneath the patch can be quantified. This scheme integrates this dynamic thermodynamic result with traditional static geometric measurements (boundary cracks, height differences) to generate a structural integrity score that considers both the surface and the interior. In this way, even if a patch appears intact in images and 3D data, structural defects underneath can be identified, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0080] In some other embodiments of this application, when the pavement patch is a newly laid dark asphalt while the original pavement is a light-colored, aged pavement, the patch may be mistakenly assigned an excessively high rate of temperature change because its heat absorption capacity is much stronger than the surrounding area. This could lead to a healthy patch being misjudged as having a high-risk defect of underlying voids. The comprehensive pavement condition assessment method provided in this application can improve the accuracy of the assessment results by calculating the grayscale difference and correcting for the rate of temperature change, thus eliminating the spurious influence of heat absorption caused by color differences.

[0081] like Figure 2 The diagram shown is another flowchart illustrating the comprehensive road surface condition assessment method provided in this application, which includes the following steps: S201. Real-time acquisition of road surface image data and 3D point cloud data, and obtaining the operation records of road repair vehicles through V2V broadcast within the work area to obtain road surface patches.

[0082] S202. Determine the boundary area between the road patch and the original road surface, obtain crack data based on the image data of the boundary area, and determine the height difference between the road patch and the original road surface based on the three-dimensional point cloud data of the boundary area.

[0083] S203. At the first preset time point and the second preset time point, collect the first thermal imaging data and the second thermal imaging data of the road patch and the original road surface, respectively.

[0084] S204. Based on the first thermal imaging data and the second thermal imaging data, calculate the patch temperature change rate of the road surface patch and the original road surface temperature change rate within a preset distance range around the road surface patch, respectively.

[0085] Steps S201-S204 and Figure 1 Steps S101-S104 in the illustrated embodiment are similar and can be found in the description of the steps; they will not be repeated here.

[0086] S205. Extract the average gray values ​​of the pavement patch area and the original pavement area from the pavement image data, and calculate the gray value difference.

[0087] Specifically, first, the registered high-resolution visible light road surface image and the vector boundaries of the road surface patch determined in S201 are retrieved. Using the patch boundaries, two analysis regions are delineated on the image: the patch region and the adjacent original road surface region.

[0088] Converting a color RGB image to a single-channel grayscale image results in each pixel in the image having only one brightness value from 0 (pure black) to 255 (pure white).

[0089] Calculate the sum of the grayscale values ​​of all pixels within the patch area, and then divide by the total number of pixels in that area to obtain the average grayscale value of the patch. Similarly, calculate the average grayscale value of the original road surface.

[0090] Finally, the absolute difference between these two average gray values ​​is calculated to obtain the gray value difference = |average gray value of patch - average gray value of original road surface|.

[0091] The theoretical basis for this step stems from fundamental physics: the color of an object determines its absorption rate of solar radiation. Typically, newly laid asphalt patches are dark black (low grayscale value), while aged, oxidized pavement is grayish-white (high grayscale value). Darker surfaces absorb more solar energy, thus naturally heating up faster under sunlight than lighter surfaces. This phenomenon is unrelated to the integrity of the structure beneath the patch, but it affects the core indicator calculated in S204—the rate of temperature change. Without considering this, a healthy patch that heats up faster simply because it's darker might be incorrectly judged as having abnormal thermal behavior. Therefore, calculating the grayscale difference is crucial for quantifying and separating the thermodynamic effects caused by differences in surface physical properties in risk assessment, improving the accuracy of pavement condition assessments for pavements with potential structural defects.

[0092] S206. In the preset emissivity correction database that maps the grayscale difference of the road surface to the surface emissivity difference, query the emissivity deviation correction value for the corresponding grayscale difference.

[0093] Specifically, the grayscale difference calculated in the previous step is used as a query index to retrieve data from an internal data table called the emissivity correction database. This database is essentially a lookup table that stores a series of mapping relationships between grayscale differences and emissivity deviation correction values.

[0094] The pre-establishment of the emissivity correction database is achieved through numerous laboratory or field calibration experiments. The process typically involves researchers selecting pavement samples with varying degrees of aging, asphalt grades, and aggregate types. Under controlled conditions, they simultaneously measure the temperature using a contact thermometer (such as a thermocouple, which can be considered the true temperature) and a thermal imager. Simultaneously, visible light images of these samples are captured, and their grayscale values ​​are calculated. By adjusting the emissivity settings of the thermal imager to match the readings of the contact thermometer, the true surface emissivity of the sample can be obtained. Repeating this process builds a large database containing the correspondence between material grayscale values ​​and their true emissivity.

[0095] S207. Correct the patch temperature change rate using the emissivity deviation correction value.

[0096] Specifically, the algorithm correction process based on the physical laws of thermal radiation is theoretically grounded in the inverse operation of the Stefan-Boltzmann law, which involves deducing the true temperature of an object's surface from the radiant energy received by the thermal imager. The simplified formula for calculating temperature using a thermal imager is proportional to the fourth power of the object's surface emissivity and the true temperature. When the emissivity is set inaccurately, the measured temperature will be inaccurate; this step aims to reverse this process.

[0097] First, the original temperature readings need to be corrected. Typically, thermal imagers have a standardized emissivity setting (e.g., the original emissivity for aged asphalt pavement). For patches that are darker and have higher emissivity, this emissivity setting is too low. Use the emissivity deviation correction value found in the previous step to calculate the patch's true emissivity: Original emissivity + Emissivity deviation correction value. Then, correct the temperatures measured at the first and second preset time points of the patch area to be closer to the true temperature using the following formula: True temperature = [(original emissivity / true emissivity) × (temperature + 273.15)^4]^(1 / 4) - 273.15.

[0098] The temperature needs to be converted to the absolute temperature scale (Kelvin) and then raised to the fourth power, before being converted back to Celsius. (Temperature + 273.15)^4 represents the radiative exitance inferred by the instrument based on the erroneous emissivity, multiplied by the correction factor (original emissivity / true emissivity).

[0099] After obtaining the corrected first-time temperature and the corrected second-time temperature respectively, the temperature change rate of the patch is recalculated, and the corrected patch temperature change rate is obtained as (corrected second-time temperature - corrected first-time temperature) / (second preset time point - first preset time point).

[0100] S208. Compare the patch temperature change rate with the original pavement temperature change rate to determine the difference value used to characterize the risk of voiding under the pavement patch.

[0101] Step S208 and Figure 1 Step S105 in the illustrated embodiment is similar and can be found in the description of the steps, which will not be repeated here.

[0102] In some embodiments, after a preliminary assessment of the overall structural risk of the pavement patch, a gridded microanalysis process targeting the interior of the patch can be further performed to generate a high-resolution risk distribution map containing spatial location information in order to achieve differentiated assessment of the location and severity of the damage.

[0103] The patch is spatially discretized. First, using the acquired pavement patch vector boundary, its internal region is divided into multiple equally sized gridded sub-regions. Essentially, this step decomposes a continuous, macroscopic analysis object into multiple discrete, microscopic analysis units.

[0104] Calculate and compare the internal temperature change rate. For each individual gridded sub-region, extract temperature information from the first and second thermal imaging data using pixels within its coverage area, and calculate its sub-region temperature change rate separately. Then, compare the temperature change rate of each sub-region with the previously calculated overall patch temperature change rate, which represents the average level. If the structure beneath the patch is uniform and healthy, the heating rates of all its sub-regions should be roughly the same. Therefore, areas whose temperature change rate significantly exceeds the overall average level are identified as anomalous sub-regions. The reason for exceeding rather than falling below is that a faster-heating area, after excluding color influences, is more likely to indicate the presence of an underlying air insulation layer (vacuum).

[0105] Quantify and locate local risks. For each identified anomalous sub-region, further calculate the difference between its sub-region temperature change rate and the original pavement temperature change rate. This difference is defined as the risk index for that anomalous sub-region. Compared to the previously calculated overall difference value, this risk index is more targeted, directly quantifying the severity of the risk at the most suspicious location within the patch. A higher risk index means that the delamination phenomenon in that local area may be more severe or cover a larger area.

[0106] A spatialized risk dataset is generated. Finally, the geospatial coordinates (or relative coordinates within their respective patches) of each anomalous sub-region are data-bound to their corresponding risk index. This operation gives the originally abstract risk values ​​a clear spatial attribute. All these coordinate-risk index data pairs are integrated together to form a dataset.

[0107] The aforementioned technical steps, through gridding of pavement patches and independent temperature change rate analysis and comparison of each sub-region, achieve a leap from macroscopic overall assessment to precise microscopic positioning. This allows for the identification and marking of unevenly distributed localized voids or early-stage defects within a seemingly homogeneous patch (these points may be masked by the overall average due to their small size). The resulting risk index set, containing spatial coordinates, provides a refined basis for maintenance decisions, thereby improving the accuracy of condition assessments for pavements with potential, localized structural defects.

[0108] S209. Surface condition scores are generated based on 3D point cloud data and image data of road surface patches.

[0109] S210. Based on crack data, height difference, and difference value, generate a structural integrity score.

[0110] Steps S209-S210 and Figure 1 Steps S106-S107 in the illustrated embodiment are similar and can be found in the description of the steps; they will not be repeated here.

[0111] In some embodiments, after identifying abnormal areas within pavement patches, a defect evolution analysis process based on graph theory analysis and time-series data comparison can be further performed to proactively identify rapidly expanding damage that poses a threat to the structure.

[0112] The identification process moves from discrete points to continuous clusters. The theoretical basis for this step is connected component analysis in graph theory. First, each anomalous sub-region data point in the previously generated dataset is abstracted as a node in a two-dimensional graph. Then, based on the spatial adjacency relationships (up, down, left, right, diagonal, etc.) between the gridded sub-regions, edges are established between adjacent nodes. By performing connectivity analysis on this constructed graph (e.g., using depth-first search or breadth-first search algorithms), all spatially connected anomalous sub-regions can be aggregated together, identifying one or more independent defect clusters.

[0113] Tracing Defects of the Same Origin Across Time. This involves accessing pre-stored historical pavement scan data and extracting historical independent defect clusters identified in the previous inspection cycle using the same method. A spatial matching process is then performed, comparing the location of each independent defect cluster identified in the current cycle with that of historical clusters. If a current cluster and a historical cluster have a significant spatial overlap, they are identified as defects of the same origin—two snapshots of the same defect at different times. This step enables the tracking of the defect lifecycle and is a prerequisite for evolutionary analysis.

[0114] The expansion trend of quantified defects is analyzed. For each identified homologous defect, the geometric centroid (i.e., the average of the coordinates of all sub-regions constituting the cluster) of its current and historical clusters is calculated. The geometric centroid can be considered as the center of gravity of the defect region. By calculating the displacement of the current geometric centroid relative to the historical geometric centroid, a two-dimensional defect expansion vector is obtained. This vector contains: the direction of the vector indicating the main expansion direction of the defect in the past period; and the magnitude (i.e., the length) of the vector, which directly quantifies the speed and magnitude of its expansion, making the description of defect development quantifiable.

[0115] Risk escalation assessment is based on evolutionary trends. When the expansion vector of a defect clearly points towards the boundary between the patch and the original pavement, and its magnitude exceeds a preset threshold (e.g., expansion exceeding 10 cm within one cycle), the pavement patch to which the defect belongs will be automatically marked as a highest-priority risk area, even if the defect's current risk index is not the highest. The logic behind this is that damage rapidly developing towards the weakest connection in the structure poses a far greater threat to the overall safety of the pavement than a static, isolated internal defect.

[0116] Generate a structural integrity score. After completing the dynamic analysis described above, return to the calculation of the structural integrity score. It first calculates the arithmetic mean of the risk indices of all abnormal sub-regions within each independent defect cluster, obtaining a cluster risk characteristic value. This characteristic value represents the average severity of the continuous defect area. Subsequently, this more representative cluster risk characteristic value replaces the previously used difference value representing the overall average level, and along with crack data, height differences, and other information, is substituted into the model to generate the final structural integrity score.

[0117] The aforementioned technical steps, by introducing graph theory analysis and time-series data comparison, quantify the temporal and spatial evolution trends of defects by calculating defect expansion vectors. This allows the identification of potential structural defects to go beyond the severity of their current state, adding two dynamic dimensions: development speed and development direction. By focusing on rapidly expanding damage in positively structurally weak links and escalating the risk accordingly, while using cluster risk characteristic values ​​that better reflect the true severity of defects to calculate scores, it is possible to identify potential defects that pose the greatest threat to long-term pavement safety, thereby improving the accuracy of pavement condition assessment for pavements with potential structural defects.

[0118] S211. Combining surface condition score and structural integrity score, output the comprehensive pavement evaluation result.

[0119] Step S211 and Figure 1 Step S108 in the illustrated embodiment is similar and can be found in the description of the steps, which will not be repeated here.

[0120] In the above embodiment, the color difference between the patch and the original pavement is quantified by calculating the grayscale difference. For example, a newly laid black asphalt patch will absorb more heat than an aged gray pavement. Next, this image feature (grayscale difference) is converted into an emissivity deviation correction value using a pre-defined database. Finally, this correction value is used to correct the original patch temperature change rate. Essentially, this process removes the pseudo-heating effect solely contributed by darker color and greater heat absorption from the observed, mixed apparent heating rate. This allows the corrected temperature change rate to more accurately reflect the heat conduction anomalies dominated by structural defects such as underlying voids, avoiding misjudging healthy patches that heat up quickly simply because of their darker color as high-risk defects. This improves the accuracy of pavement condition assessment for pavements with potential structural defects.

[0121] The following describes an exemplary road surface condition comprehensive assessment system 300 provided in an embodiment of this application. Figure 3 This is an exemplary hardware structure diagram of the road surface condition comprehensive assessment system 300 provided in this application embodiment.

[0122] In some embodiments, the road surface condition comprehensive assessment system 300 includes a computer device. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores data. The network interface communicates with other external terminals or servers via a network connection. In some embodiments, the network interface can be a wired network interface; in some embodiments, it can also be a wireless network interface. When the computer program is executed by the processor, it implements the methods described in the embodiments of this application.

[0123] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0124] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0125] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0126] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0127] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A comprehensive assessment method for road surface condition, characterized in that, include: The road surface scanning data, which includes road surface image data and 3D point cloud data, is collected in real time, and the operation records of the road repair engineering vehicle are obtained through V2V broadcast on the intranet of the work area to obtain road surface patches. The boundary region between the road patch and the original road surface is determined, crack data is obtained based on the image data of the boundary region, and the height difference between the road patch and the original road surface is determined based on the three-dimensional point cloud data of the boundary region. At a first preset time point and a second preset time point, first thermal imaging data and second thermal imaging data of the road patch and the original road surface are collected respectively. Based on the first thermal imaging data and the second thermal imaging data, the patch temperature change rate of the road patch and the original road surface temperature change rate within a preset distance range around the road patch are calculated respectively. The temperature change rate of the patch is compared with the temperature change rate of the original pavement to determine the difference value used to characterize the risk of voiding under the pavement patch. A surface condition score is generated based on the three-dimensional point cloud data and image data of the road surface patch. Based on the crack data, the height difference, and the difference value, a structural integrity score is generated; By combining the surface condition score and the structural integrity score, a comprehensive pavement evaluation result is output.

2. The method according to claim 1, characterized in that, Before comparing the patch temperature change rate with the original pavement temperature change rate to determine the difference value used to characterize the risk of delamination beneath the pavement patch, the method further includes: From the road surface image data, the average gray values ​​of the road surface patch area and the original road surface area are extracted respectively, and the gray value difference is calculated. In the preset emissivity correction database that maps the grayscale difference of road surface to the surface emissivity difference, query the emissivity deviation correction value corresponding to the grayscale difference; The emissivity deviation correction value is used to correct the patch temperature change rate.

3. The method according to claim 1, characterized in that, After comparing the patch temperature change rate with the original pavement temperature change rate to determine a difference value characterizing the risk of delamination beneath the pavement patch, the method further includes: The road surface patch is divided into multiple gridded sub-regions of preset size; Based on the first thermal imaging data and the second thermal imaging data, the temperature change rate of each gridded sub-region is calculated separately. The temperature change rate of each sub-region is compared with the overall patch temperature change rate of the road surface patch to identify abnormal sub-regions; the abnormal sub-regions are gridded sub-regions whose temperature change rate exceeds the overall patch temperature change rate. Calculate the difference between the sub-region temperature change rate and the original road surface temperature change rate for each of the abnormal sub-regions, and define the difference as the risk index for the corresponding abnormal sub-region. The spatial coordinates of each of the abnormal sub-regions are associated with the corresponding risk index to generate a dataset containing data points of the abnormal sub-regions.

4. The method according to claim 3, characterized in that, The process of generating a structural integrity score based on the crack data, the height difference, and the difference value specifically includes: Each data point in the dataset is considered a graph node in two-dimensional space. Based on the spatial adjacency relationship of the gridded sub-regions, connectivity analysis is performed on all the graph nodes to identify one or more independent defect clusters composed of spatially continuous anomalous sub-regions. Calculate the average risk index of all abnormal sub-regions within each independent defect cluster to obtain a cluster risk characteristic value that characterizes the overall risk level of the independent defect cluster. Based on the crack data, the height difference, and the cluster risk characteristic value, a structural integrity score is generated.

5. The method according to claim 4, characterized in that, After identifying one or more independent defect clusters consisting of spatially contiguous anomalous sub-regions, the method further includes: Extract the historical independent defect clusters from the historical road surface scanning data from the previous evaluation period; Each currently identified independent defect cluster is spatially matched with the historical independent defect clusters to identify defects of the same origin. Calculate the geometric centroid of each independent defect cluster and the historical independent defect cluster of each of the aforementioned homologous defects to obtain the current geometric centroid and the historical geometric centroid. Calculate the displacement of the current geometric centroid relative to the historical geometric centroid to obtain the defect expansion vector that characterizes the evolution trend of the corresponding defect in time and space; When the direction of the defect propagation vector points to the boundary region of the road patch and the magnitude of the defect propagation vector exceeds a preset propagation length threshold, the road patch is marked as a high-priority risk area.

6. The method according to claim 1, characterized in that, Before calculating the patch temperature change rate of the road surface patch and the original road surface temperature change rate within a preset distance range around the road surface patch based on the first thermal imaging data and the second thermal imaging data, the method further includes: From the three-dimensional point cloud data, extract the corresponding patch laser reflection intensity and the original road surface laser reflection intensity for the road surface patch and the original road surface, respectively; The laser reflection intensity of the patch, the laser reflection intensity of the original road surface, and the preset dry and clean road surface reflection intensity benchmark value are compared to obtain the wet area and the reflection deviation value corresponding to each wet area; the wet area is the area where the corresponding reflection deviation value is greater than the preset deviation value threshold. Multiply the reflection deviation value by a preset temperature compensation coefficient to obtain a temperature correction amount, and add the temperature correction amount to the corresponding humid area to generate corrected first thermal imaging data and second thermal imaging data.

7. The method according to claim 6, characterized in that, The step of multiplying the reflection deviation value by a preset temperature compensation coefficient to obtain a temperature correction amount, and adding the temperature correction amount to the corresponding humid area to generate corrected first thermal imaging data and second thermal imaging data, specifically includes: Within a predetermined proximity range of each of the wetted areas, a reference area is determined where the laser reflection intensity is higher than the reference value of the dry and clean road surface reflection intensity. Based on the first thermal imaging data, calculate the first temperature difference between the humid region and the reference region; Based on the second thermal imaging data, a second temperature difference between the humid region and the reference region is calculated; The difference between the first temperature difference and the second temperature difference is used to obtain the dynamic evaporative cooling correction value; The corrected second thermal imaging data is obtained by adding the dynamic evaporative cooling correction value and the second thermal imaging data of the humid area.

8. A comprehensive road surface condition assessment system, characterized in that, The road surface condition comprehensive assessment system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the road surface condition comprehensive assessment system to perform the method as described in any one of claims 1-7.

9. A computer program product containing instructions, characterized in that, When the computer program product is run on the road condition comprehensive assessment system, the road condition comprehensive assessment system performs the method as described in any one of claims 1-7.

10. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is run on the road condition comprehensive assessment system, the road condition comprehensive assessment system performs the method as described in any one of claims 1-7.

Images