Asphalt pavement visual blackness evaluation method and system based on human eye visual simulation

Abstract

The invention relates to an asphalt pavement visual blackness evaluation method and system based on human eye visual simulation, and the method comprises the steps: collecting apparent optical parameters of a monitoring point, inputting the apparent optical parameters into an asphalt pavement visual blackness comprehensive evaluation model, and obtaining a blackness comprehensive score; the asphalt pavement visual blackness comprehensive evaluation model is obtained by training a blackness perception intensity model and a color deviation perception intensity model by using a training set; the training set comprises original optical parameter data; in the training process, the maximum value of the Spearman level correlation coefficient between the output value of the blackness comprehensive evaluation model and the subjective score is taken as an optimization target, and the coefficients in the model are iteratively optimized by adopting an optimization algorithm to determine the final model coefficient. According to the method, the tiny attenuation of the apparent characteristics of the pavement can be identified before the pavement is subjected to structural damage, and early warning and preventive maintenance support are realized. Meanwhile, evaluation thresholds are self-defined according to road grades, regional climate and maintenance standards, and different materials and regional road conditions are adapted.

Description

Technical Field

[0001] This invention relates to the field of pavement evaluation technology, and in particular to a method and system for evaluating the visual density of asphalt pavement based on human visual simulation. Background Technology

[0002] Real-time pavement monitoring and maintenance decisions are crucial for ensuring driving safety and extending the service life of asphalt pavements. Currently, mainstream testing methods mainly include three categories: laboratory material performance testing, intelligent non-destructive testing technology, and pavement structure performance evaluation. Among them, laboratory material performance testing methods, while providing accurate data, have limitations such as damaging the pavement during sampling, long testing cycles, and difficulty in achieving large-area rapid testing. Intelligent testing technologies (such as laser scanning and hyperspectral imaging), which have been vigorously developed in recent years, have the advantages of non-destructive testing, but they generally rely on expensive equipment, require highly specialized operation, and have limited environmental adaptability, making it difficult to widely promote and apply them in routine maintenance. As for pavement structure performance evaluation methods based on fixed template devices, they have significant limitations because they cannot accurately reflect the pavement condition under the coupled effects of actual traffic loads and the environment.

[0003] In recent years, detection methods based on surface features have attracted attention due to their advantages such as high efficiency, economy, and ease of real-time monitoring. Existing technologies disclose a machine vision-based method for detecting road surface quality, which can improve the accuracy of crack detection by analyzing the changes in brightness and temperature of road cracks before and after watering. Existing technologies also disclose a method and system for rapid road quality detection, which can reduce the impact of road shadows on the accuracy of crack detection. These technologies have promoted the development of automated detection of road surface conditions, but their core focus remains on the identification and classification of existing defects (such as cracks), which is a form of "post-diagnosis" and cannot effectively warn of early performance degradation of road materials. In fact, macroscopic defects such as cracks are usually late manifestations of intrinsic material performance degradation such as asphalt aging and bond failure. Studies have shown that there is a significant correlation between the surface optical properties of asphalt pavement (such as blackness, yellowness, and reflectivity) and key intrinsic quality indicators such as asphalt film thickness, aging degree, and skid resistance. These changes often precede the appearance of macroscopic defects and can therefore serve as sensitive indicators of early pavement performance degradation.

[0004] However, current systematic research on the correlation between asphalt pavement blackness and service performance, both domestically and internationally, is still insufficient. In particular, there is a lack of mature methods based on human visual perception mechanisms, mapping multi-source optical parameters to perceived blackness intensity, and establishing a comprehensive evaluation and early diagnosis method for pavement visual blackness. Therefore, developing an asphalt pavement visual blackness evaluation method that can simulate human vision and balance accuracy, efficiency, and economy is of great significance for overcoming existing detection bottlenecks and promoting the development of preventative maintenance technologies. Summary of the Invention

[0005] To address the problems existing in the prior art, the purpose of this invention is to provide a method and system for evaluating the visual blackness of asphalt pavement based on human visual simulation. By simulating the human visual perception mechanism, a comprehensive evaluation model for the visual blackness of asphalt pavement is established, enabling an economical, efficient, and accurate evaluation of pavement color indicators based on apparent optical characteristics.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A method for evaluating the visual blackness of asphalt pavement based on human visual simulation includes:

[0008] The apparent optical parameters of the monitoring points are collected and input into the asphalt pavement visual blackness comprehensive evaluation model to obtain the comprehensive blackness score; the asphalt pavement visual blackness comprehensive evaluation model is obtained by training a blackness perception intensity model and a color deviation perception intensity model using a training set; the training set includes: original optical parameter data;

[0009] During the training process, the maximum value of the Spearman rank correlation coefficient between the output value of the comprehensive blackness evaluation model and the subjective score is taken as the optimization objective. An optimization algorithm is used to iteratively optimize the coefficients in the model to determine the final model coefficients.

[0010] Optionally, the apparent optical parameters collected at the monitoring points include:

[0011] The area of ​​the asphalt pavement to be tested was divided into multiple test zones and test points were set up. Non-reflective markers were used to calibrate the center of the test points and to collect the surface optical parameters of each monitoring point.

[0012] During the collection of the apparent scientific parameters, it is determined whether the standard deviation of the measured values ​​of the same measurement point in the same measurement area is less than the first target threshold. If the standard deviation is less than the first target threshold, the measurement is redone. It is also determined whether the coefficient of variation of the data for the entire road segment exceeds the second target threshold. If the coefficient of variation of the data for the entire road segment exceeds the second target threshold, the road surface uniformity or measurement error is analyzed.

[0013] Optionally, the blackness perception intensity model includes:

[0014] ;

[0015] in, , , These are the theoretical maximum values ​​or calibration upper limits for each optical parameter. , The chromaticity coordinate reference value is the ideal value for a fresh asphalt pavement. , , , These are the contribution weighting coefficients for each corresponding optical parameter. It is an exponentially decaying function. For brightness and darkness, For red and green levels, The degree of yellow-blue, For reflectivity, For blackness.

[0016] Optionally, the color deviation perception intensity model includes:

[0017] ;

[0018] in, , , These are the model weight coefficients. It is a nonlinear factor. Yellowness, d This is the hue shift value. For red and green levels, The degree of yellow-blue, , The reference value for the chromaticity coordinates of an ideal fresh asphalt pavement.

[0019] Optionally, the comprehensive evaluation model for visual blackness of asphalt pavement includes:

[0020] ;

[0021] in, This is a smoothing constant used for adjustment. The influence intensity on the overall blackness score, where c is the optimization coefficient, so that the final result is within the range of 0 to 100.

[0022] Optionally, determining the contribution weight coefficients of the blackness perception intensity model in the final model coefficients includes:

[0023] ;

[0024] in, , , These are the theoretical maximum values ​​or calibration upper limits of each optical parameter.

[0025] Optionally, obtaining the maximum value of the Spearman rank correlation coefficient includes:

[0026] ;

[0027] in, For the first one sample The difference between the ranking and the S ranking, where N is the total number of samples.

[0028] To achieve the above objectives, the present invention also provides an asphalt pavement visual blackness evaluation system based on human eye vision simulation, comprising:

[0029] The comprehensive blackness evaluation module is used to collect the apparent optical parameters of the monitoring points, input the apparent optical parameters into the asphalt pavement visual blackness comprehensive evaluation model, and obtain the comprehensive blackness score; the asphalt pavement visual blackness comprehensive evaluation model is obtained by training a blackness perception intensity model and a color deviation perception intensity model using a training set; the training set includes: original optical parameter data;

[0030] The comprehensive blackness evaluation module also includes:

[0031] The model training submodule is used to optimize the coefficients in the model during training by using the maximum value of the Spearman rank correlation coefficient between the output value of the comprehensive blackness evaluation model and the subjective score as the optimization objective, and to determine the final model coefficients by using an optimization algorithm to iteratively optimize the coefficients in the model.

[0032] The beneficial effects of this invention are as follows:

[0033] This invention overcomes the limitations of traditional methods that rely solely on physical instrument data or pure human experience, constructing a visual perception parameter mapping model centered on Blackness Perceived Intensity (BPI) and Color Deviation Perceived Intensity (SPI). This model cleverly simulates the human eye's sensitivity to road surface blackness and color purity, as well as the "threshold effect," ensuring that objective optical measurement data accurately reflects the subjective visual perception of the human eye. Finally, by maximizing the Spearman correlation coefficient, the collective subjective evaluation wisdom of senior engineers is solidified into the optimization algorithm, ensuring that the evaluation results are both objective and reliable, and consistent with the subjective cognition of industry experts. This solves the technical challenges of highly subjective purely manual evaluations and the disconnect between purely instrumental evaluations and real visual perception.

[0034] This invention enables timely and non-destructive diagnosis of early performance degradation in asphalt pavements, demonstrating significant early warning capabilities. Compared to traditional detection methods that rely on visible defects such as cracks and potholes, this invention keenly identifies changes in asphalt pavement opacity as a leading indicator of intrinsic quality degradation, such as aging and film thinning. Through systematic optical parameter acquisition and simulation analysis, it can detect subtle attenuations in the apparent characteristics of asphalt pavements before structural damage occurs, thus providing an early warning of pavement quality degradation and offering a valuable time window for implementing cost-effective preventive maintenance. This overcomes the significant drawback of existing image detection methods, which suffer from evaluation lag. Furthermore, this method utilizes a portable optical inspection instrument consisting of a light source, detector array, beam splitter, and data processing unit. It possesses the capability for non-destructive testing of road pavement optical parameters, requiring no destructive sampling. The operation is simple and efficient, making it suitable for rapid screening and long-term monitoring of a wide range of pavement conditions.

[0035] This invention integrates a visual blackness evaluation method for asphalt pavement based on human visual simulation into a software system, achieving closed-loop management of the entire process from data acquisition, processing, calculation to result output. The system's output of a comprehensive blackness score is intuitive and easy to use, directly supporting maintenance decisions. The system supports customizing evaluation thresholds based on road grade, regional climate, and maintenance standards, effectively adapting to pavement conditions of different regions and materials, significantly improving the method's universality and result reliability. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a flowchart of a method for evaluating the visual blackness of asphalt pavement based on human visual simulation, according to an embodiment of the present invention. Detailed Implementation

[0038] 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.

[0039] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0040] like Figure 1 As shown in the figure, this embodiment discloses a method for evaluating the visual blackness of asphalt pavement based on human visual simulation, including: collecting the apparent optical parameters of monitoring points, inputting the apparent optical parameters into the comprehensive evaluation model of visual blackness of asphalt pavement, and obtaining the comprehensive blackness score; the comprehensive evaluation model of visual blackness of asphalt pavement is obtained by training a blackness perception intensity model and a color deviation perception intensity model using a training set; the training set includes: original optical parameter data; during the training process, taking the maximum value of the Spearman rank correlation coefficient between the output value of the comprehensive blackness evaluation model and the subjective score as the optimization objective, the optimization algorithm is used to iteratively optimize the coefficients in the model to determine the final model coefficients.

[0041] Furthermore, the apparent optical parameters of the monitoring points are collected as follows: the area of ​​the asphalt pavement to be tested is divided into multiple test zones and test points are set up. Non-reflective markers are used to calibrate the center of the test points for collecting the apparent optical parameters of each monitoring point. During the collection of apparent optical parameters, it is determined whether the standard deviation of the measured values ​​of the test points in the same test zone is less than the first target threshold. If the standard deviation is less than the first target threshold, the test points are re-set up. It is also determined whether the coefficient of variation of the data for the entire road section exceeds the second target threshold. If the coefficient of variation of the data for the entire road section exceeds the second target threshold, the pavement uniformity or measurement error is analyzed.

[0042] Specifically, the planning of asphalt pavement measuring points involves dividing and setting up measuring points in the asphalt pavement area to be measured, excluding traffic markings, defective areas, color-marked areas, and road ancillary facilities based on the condition of each measuring point, and the remaining measuring points constitute the measuring point set.

[0043] In this step, the road surface temperature must be controlled within the range of 0°C to 70°C. If there are obvious water stains, oil stains, or chemical residues (such as de-icing agents) on the road surface before measurement, it must be cleaned and allowed to dry completely before measurement under natural, uniform lighting to avoid moisture affecting optical reflectivity measurement and interference from strong light reflection or shadows. During measurement, each 100-meter section is divided into a measurement area (section), and each section is further divided into three measurement domains (blocks). Each measurement domain (1m × 1m square) is divided into an average of 16 measurement points (4×4 grid). Non-reflective markers (such as chalk) are used to lay out the measurement points on the asphalt pavement area to be measured. If there are areas of different colors or conditions on the road surface, the largest area should be selected as the main measurement target, avoiding other colors or abnormal areas.

[0044] Optical parameter acquisition: The surface optical parameters of each measuring point were collected using a portable optical detector. After the validity of the collected data was verified, an optical database for asphalt pavement was established. The optical parameters include black value (MY), yellowness (YI), surface reflectance (LRV), hue deviation value (dM), brightness value (L), red-green value (a*), and yellow-blue value (b*), as shown in Table 1.

[0045] Table 1. Meaning of Asphalt Pavement Opacity Database Data

[0046]

[0047] All the above data can be obtained directly using a colorimeter. Before measurement, the colorimeter needs to be calibrated using a standard blackboard (reflectivity ≤5%) and a whiteboard (reflectivity >90%). The method of using the colorimeter is to place the colorimeter probe vertically against the asphalt pavement, ensuring there are no gaps or tilts, and apply constant pressure at 3 seconds per measurement. Each measuring point should be measured 3 times consecutively, with an interval of 10 seconds between each measurement. Obvious outliers (single value deviating from the mean ±10%) should be eliminated, and the arithmetic mean of the remaining data should be taken as the final result. If the difference between the measured values ​​of adjacent measuring points is >15%, the pavement should be cleaned and the measurement repeated. If the remeasurement result is still abnormal, the area should be marked and the possible causes recorded.

[0048] Furthermore, visual perception parameter mapping: the collected optical parameters are converted into human eye visual sensitivity parameters; visual sensitivity parameters include blackness perception intensity (HPI) and color deviation perception intensity (SPI).

[0049] In this step, the calculation model for perceived blackness intensity (BPI) is as follows:

[0050] ;

[0051] in, , , These are the theoretical maximum values ​​or calibration upper limits for each optical parameter. , The chromaticity coordinate reference value is the ideal value for a fresh asphalt pavement. , , , These are the contribution weighting coefficients for each corresponding optical parameter. It is an exponentially decaying function.

[0052] The calculation model for color cast perceived intensity (SPI) is as follows:

[0053]

[0054] in, , , These are the model weight coefficients. These are nonlinear factors, all determined through optimization algorithms. This constrains the theoretical output range of BPI to [0,100].

[0055] A higher BPI value indicates that the road surface appears darker and purer to the human eye. In this model... It is an exponential decay function, simulating the sensitivity of the human eye to color purity;

[0056] Furthermore, the larger the SPI value, the more severe the degree of road surface discoloration, pollution, and aging (yellowing, whitishness) perceived by the human eye, and the worse the road surface quality; the γ in this model simulates the "threshold effect" of human eye's sensitivity to pollution and discoloration.

[0057] Constructing a comprehensive evaluation model for visual blackness of asphalt pavement: Establishing a comprehensive evaluation model for visual blackness of asphalt pavement with BPI and SPI as core input variables, and outputting a comprehensive blackness score G;

[0058] In this step, the expression for the comprehensive evaluation model of visual blackness of asphalt pavement is: ;in, This is a smoothing constant used for adjustment. The influence intensity on the overall blackness score, where c is the optimization coefficient, so that the final result is within the range of 0 to 100.

[0059] When the road surface color is pure and there is no color deviation, the SPI approaches 0, the adjustment factor approaches 1, and the overall blackness score G≈BPI is obtained.

[0060] When there is severe color deviation or pollution on the road surface, the SPI is large, the adjustment factor tends to 0, and the score G will be significantly reduced.

[0061] Model coefficient calibration and adaptive optimization: Based on a large amount of sample data, the model coefficients are calibrated and optimized. The model coefficient calibration and adaptive optimization method is as follows:

[0062] N sets of optical parameter data of pavement under different aging conditions were collected as training set; an evaluation team composed of several senior maintenance engineers conducted blind evaluation of the above N sets of pavement samples and gave subjective blackness scores S; with the optimization objective of maximizing the Spearman rank correlation coefficient between the model output value G and the subjective score S, an intelligent optimization algorithm was used to iteratively optimize the coefficients in the model to determine the final model parameters.

[0063] ;

[0064] in, For the first one sample The difference between the ranking and the S ranking, where N is the total number of samples; The closer the value is to 1, the more consistent the blackness ranking given by the simulation system is with the actual road surface blackness ranking. In other words, the system can effectively simulate human vision and make an accurate evaluation of the visual blackness of the road surface.

[0065] The coefficients that need to be optimized include: the weighting coefficients in the BPI model. , , , Weight coefficients in the SPI model , , And the nonlinear factor γ, and the smoothing constant k in the comprehensive blackness evaluation model.

[0066] Visual blackness output of asphalt pavement: The optical parameters of the pavement to be tested are input into the calibrated and optimized evaluation model to calculate the comprehensive blackness score G.

[0067] The threshold range can be customized within the system based on different road grades, regional climate characteristics, and maintenance standards. Specifically, the adjustment needs to take into account the following factors:

[0068] Higher-grade roads (such as expressways and trunk roads) should adopt stricter thresholds to implement higher standards of preventive maintenance; in areas with harsh environments such as rain, heat, and strong ultraviolet radiation, the lower limit of the threshold for each grade should be appropriately lowered to trigger maintenance warnings in advance; in areas with sufficient maintenance budgets and proactive strategies, more sensitive thresholds can be adopted to promote early maintenance intervention.

[0069] Based on the above technical solution, the present invention provides a specific embodiment as shown below. This part aims to describe in detail the specific implementation of a method for evaluating the visual blackness of asphalt pavement based on human eye vision simulation.

[0070] Since the asphalt pavement visual blackness evaluation method based on human eye vision simulation provided by this invention mainly involves the planning of asphalt pavement measurement points on actual pavement, this embodiment specifically includes: optical parameter acquisition, visual perception parameter mapping and blackness comprehensive evaluation model construction, model coefficient calibration and adaptive optimization, and asphalt pavement visual blackness output.

[0071] Since the implementation of this invention requires a large amount of sample data to calibrate and optimize the model coefficients, hundreds of samples are typically needed in practical applications to ensure the model's universality and accuracy. This embodiment, for the purpose of clearly demonstrating the implementation process and method of this invention and facilitating understanding, only selects 10 measurement points for illustrative demonstration and explanation. This data scale is merely illustrative and does not represent the sample size required for actual applications. The specific steps are as follows:

[0072] Optical parameter acquisition and establishment of an optical parameter database for asphalt pavement;

[0073] Taking a field survey of a newly commissioned Class I highway as an example, the asphalt pavement was first divided and laid out according to the measurement point set establishment method proposed in this invention. Then, a portable optical detector was used to collect the optical parameters of the asphalt pavement from the measurement point set. The validity of the data was verified and processed, and an optical parameter database was established. Table 2 shows the optical parameters of some measurement points as examples.

[0074] Table 2 Basic Data of Optical Database for Asphalt Pavement Measurement Points

[0075]

[0076] Visual perception parameter mapping converts the collected optical parameters into human visual sensitivity parameters and constructs a comprehensive evaluation model for the visual blackness of asphalt pavement.

[0077] Through on-site measurements at various measuring points along the road section, the calibrated upper limits of various optical parameters under the colorimeter, as well as the chromaticity coordinate reference values ​​of fresh asphalt pavement, were obtained. , As shown in Table 3. Further, the collected optical parameters are input into the BIP and SIP calculation models, and a set of exemplary model coefficients are preset to convert the optical parameters of each measurement point into human visual sensitivity parameters.

[0078] Table 3 Upper Limit and Reference Values ​​of Optical Parameters

[0079]

[0080] Based on this, according to the calculation models of visual blackness perception intensity (BPI) and color shift perception intensity (SPI) of asphalt pavement proposed in this invention, the BPI and SPI of each measuring point are calculated.

[0081] Constructing a comprehensive evaluation model for the visual blackness of asphalt pavement:

[0082] By setting a smoothing constant, the comprehensive blackness score of each measuring point is obtained based on the comprehensive evaluation model G of visual blackness of asphalt pavement with BPI and SPI as the core input variables.

[0083] ;

[0084] Where k is a smoothing constant used to adjust the intensity of the influence of SPI on the overall blackness score.

[0085] Asphalt pavement quality grading and output: Substitute the model coefficients in Table 4 into the asphalt pavement visual blackness comprehensive evaluation model, and calculate the comprehensive blackness score G of all measuring points (see Table 5).

[0086] Table 4 Optimal Model Coefficient Combination

[0087]

[0088] Table 5. Overall Emissivity Scores of Asphalt Pavement at Various Measuring Points

[0089]

[0090] The results show that the method of this invention can efficiently and accurately quantify the appearance condition of asphalt pavement. The evaluation scores of all measuring points are distributed between 80 and 95 points, indicating that the asphalt pavement has good blackness. Except for a few areas with quality scores below 80 points that are included in the mid-term maintenance plan, only routine preventative maintenance is needed. Furthermore, the evaluation results are highly consistent with the pavement blackness, color deviation, and other appearance characteristics reflected by the optical parameters of the measuring points, proving the rationality and effectiveness of the model calculation. In summary, this invention has advantages such as simple operation, objective evaluation, and intuitive results, providing a reliable technical means for rapid, non-destructive testing of the visual blackness of asphalt pavement and scientific maintenance decision-making.

[0091] This embodiment also provides an asphalt pavement visual blackness evaluation system based on human eye vision simulation, including: a blackness comprehensive evaluation module, used to collect the apparent optical parameters of monitoring points, input the apparent optical parameters into the asphalt pavement visual blackness comprehensive evaluation model, and obtain a comprehensive blackness score; the asphalt pavement visual blackness comprehensive evaluation model is obtained by training a blackness perception intensity model and a color deviation perception intensity model using a training set; the training set includes: original optical parameter data; the blackness comprehensive evaluation module also includes:

[0092] The model training submodule is used to optimize the coefficients in the model during training by using the maximum value of the Spearman rank correlation coefficient between the output value of the comprehensive blackness evaluation model and the subjective score as the optimization objective, and to determine the final model coefficients by using an optimization algorithm to iteratively optimize the coefficients in the model.

[0093] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for evaluating the visual blackness of asphalt pavement based on human visual simulation, characterized in that, include: Collect the apparent optical parameters of the monitoring points, input the apparent optical parameters into the asphalt pavement visual blackness comprehensive evaluation model, and obtain the comprehensive blackness score; The comprehensive evaluation model for visual blackness of asphalt pavement is obtained by training a blackness perception intensity model and a color deviation perception intensity model using a training set; the training set includes: original optical parameter data; During the training process, the maximum value of the Spearman rank correlation coefficient between the output value of the comprehensive blackness evaluation model and the subjective score is taken as the optimization objective. An optimization algorithm is used to iteratively optimize the coefficients in the model to determine the final model coefficients.

2. The method for evaluating the visual blackness of asphalt pavement based on human visual simulation according to claim 1, characterized in that, The apparent optical parameters collected at the monitoring points include: The area of ​​the asphalt pavement to be tested was divided into multiple test zones and test points were set up. Non-reflective markers were used to calibrate the center of the test points and to collect the surface optical parameters of each monitoring point. During the collection of the apparent scientific parameters, it is determined whether the standard deviation of the measured values ​​of the same measurement point in the same measurement area is less than the first target threshold. If the standard deviation is less than the first target threshold, the measurement is redone. It is also determined whether the coefficient of variation of the data for the entire road segment exceeds the second target threshold. If the coefficient of variation of the data for the entire road segment exceeds the second target threshold, the road surface uniformity or measurement error is analyzed.

3. The method for evaluating the visual blackness of asphalt pavement based on human visual simulation according to claim 1, characterized in that, The perceived blackness intensity model includes: ； in, , , These are the theoretical maximum values ​​or calibration upper limits for each optical parameter. , The chromaticity coordinate reference value is the ideal value for a fresh asphalt pavement. , , , These are the contribution weighting coefficients for each corresponding optical parameter. It is an exponentially decaying function. For brightness and darkness, For red and green levels, The degree of yellow-blue, For reflectivity, For blackness.

4. The method for evaluating the visual blackness of asphalt pavement based on human visual simulation according to claim 1, characterized in that, The color deviation perception intensity model includes: ； in, , , These are the model weight coefficients. It is a nonlinear factor. Yellowness, d This is the hue shift value. For red and green levels, The degree of yellow-blue, , The reference value for the chromaticity coordinates of an ideal fresh asphalt pavement.

5. The method for evaluating the visual blackness of asphalt pavement based on human visual simulation according to claim 1, characterized in that, The comprehensive evaluation model for visual blackness of asphalt pavement includes: ； in, This is a smoothing constant used for adjustment. The influence intensity on the overall blackness score, where c is the optimization coefficient, so that the final result is within the range of 0 to 100.

6. The method for evaluating the visual blackness of asphalt pavement based on human visual simulation according to claim 3, characterized in that, The contribution weighting coefficients of the blackness perception intensity model in the final model coefficients include: ； in, , , These are the theoretical maximum values ​​or calibration upper limits of each optical parameter.

7. The method for evaluating the visual blackness of asphalt pavement based on human visual simulation according to claim 1, characterized in that, Obtaining the maximum value of the Spearman rank correlation coefficient includes: ； in, For the first one sample The difference between the ranking and the S ranking, where N is the total number of samples.

8. A visual blackness evaluation system for asphalt pavement based on human visual simulation, implemented by the method according to any one of claims 1-7, characterized in that, include: The comprehensive blackness evaluation module is used to collect the apparent optical parameters of the monitoring points, input the apparent optical parameters into the visual blackness comprehensive evaluation model of asphalt pavement, and obtain the comprehensive blackness score. The comprehensive evaluation model for visual blackness of asphalt pavement is obtained by training a blackness perception intensity model and a color deviation perception intensity model using a training set; the training set includes: original optical parameter data; The comprehensive blackness evaluation module also includes: The model training submodule is used to optimize the coefficients in the model during training by using the maximum value of the Spearman rank correlation coefficient between the output value of the comprehensive blackness evaluation model and the subjective score as the optimization objective, and to determine the final model coefficients by using an optimization algorithm to iteratively optimize the coefficients in the model.

Images