Road surface disease development prediction method, device, system, equipment and medium

Abstract

The application provides a pavement disease development prediction method, device, system, equipment and medium, and relates to the technical field of highway maintenance.The method comprises the following steps: acquiring a pavement image of a to-be-predicted road section as a target pavement image; performing pavement disease identification on the target pavement image to acquire type information and geometric information of pavement diseases in the target pavement image as a target pavement disease identification result; and inputting the target pavement disease identification result into a pavement disease development prediction model to acquire a pavement disease development prediction result of the to-be-predicted road section output by the pavement disease development prediction model.The pavement disease development prediction method, device, system, equipment and medium provided by the application can more accurately and efficiently predict the development trend of microscopic pavement diseases, can provide decision support for preventive maintenance timing, can increase the service life of roads, and can improve the operation efficiency of roads.

Description

Technical Field

[0001] This invention relates to the field of highway maintenance technology, and in particular to a method, device, system, equipment and medium for predicting the development of pavement defects. Background Technology

[0002] Preventive road maintenance refers to a series of measures taken before obvious road surface defects appear to maintain and extend the service life of roads and reduce the occurrence and spread of defects. Preventive maintenance can effectively reduce maintenance costs, reduce traffic hazards, and improve road safety and comfort, and has gradually become a consensus among road maintenance practitioners.

[0003] Microscopic pavement distress refers to minor, localized damage or deformation on the pavement surface. Severe pavement distress usually develops from microscopic pavement distress and poses a significant threat to various pavement performance characteristics. Therefore, accurately predicting the development trend of microscopic pavement distress is of great importance for preventive road maintenance.

[0004] However, traditional related technologies struggle to accurately predict the development trend of micro-pavement defects. Therefore, how to more accurately predict the development trend of micro-pavement defects is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This invention provides a method, device, system, equipment, and medium for predicting the development of pavement distress, in order to overcome the shortcomings of existing technologies that make it difficult to accurately predict the development trend of micro-pavement distress, and to achieve more accurate prediction of the development trend of micro-pavement distress.

[0006] This invention provides a method for predicting the development of pavement defects, comprising:

[0007] Obtain road surface images of the road segment to be predicted, and use them as target road surface images;

[0008] The target pavement image is subjected to pavement distress identification to obtain the type and geometric information of the pavement distress in the target pavement image, which is used as the target pavement distress identification result.

[0009] The target pavement distress identification results are input into the pavement distress development prediction model to obtain the pavement distress development prediction results of the road segment to be predicted output by the pavement distress development prediction model.

[0010] The pavement distress development prediction results include: time-series data on the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the pavement time-series images of the first sample road segment and the sample pavement distress identification time-series results of the pavement time-series images; the pavement time-series images include pavement images of the first sample road segment collected at different times and arranged in chronological order; the sample pavement distress identification time-series results include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distresses as well as the type, location, and geometric information of the pavement distresses.

[0011] According to the present invention, a method for predicting the development of pavement defects is provided, wherein the pavement defect development prediction model includes: an identification module, a prediction module, and an output module;

[0012] The prediction module is constructed based on the following steps:

[0013] Based on the time-series results of pavement distress identification in the sample, the time required for each type of pavement distress to change from its appearance to its severity becoming moderate is obtained.

[0014] The prediction module is constructed based on the time required for each type of pavement distress to change from its appearance to a moderate severity.

[0015] The correspondence between the types of pavement defects and the evaluation parameters and severity classification thresholds is predefined, and the severity classification thresholds include a first threshold for classifying mild and moderate defects.

[0016] According to the method for predicting the development of pavement distress provided by the present invention, the step of obtaining the time required for each type of pavement distress to change from its appearance to its severity becoming moderate based on the time series results of the sample pavement distress identification includes:

[0017] Based on the location information of pavement defects in the pavement images of each first sample road segment, the pavement defects in the pavement images of each first sample road segment are aligned with entities, and the same pavement defects are associated in the pavement images of each first sample road segment based on the entity alignment results.

[0018] For each identical pavement defect, based on the geometric and type information of each identical pavement defect in the pavement images of different first sample road segments, the pavement images of the first sample road segments whose severity first becomes moderate are determined.

[0019] Based on the acquisition time of the first sample road segment of each identical pavement defect when it first appears, and the acquisition time of the first sample road segment of each identical pavement defect when its severity first changes to moderate, the time required for each identical pavement defect to change from its appearance to its severity becoming moderate is obtained.

[0020] Based on the time required for each identical pavement distress to change from severity to moderate from its appearance, the time required for each type of pavement distress to change from severity to moderate from its appearance is obtained.

[0021] According to a method for predicting pavement distress development provided by the present invention, the step of inputting the target pavement distress identification result into a pavement distress development prediction model and obtaining the pavement distress development prediction result of the road segment to be predicted output by the pavement distress development prediction model includes:

[0022] The target pavement defect identification results are input into the identification module. Based on these results, the identification module determines the number of pavement defects in the target pavement image with a severity level of not less than moderate. The number of pavement defects with a severity level of not less than moderate in the target pavement image, output by the identification module, is then used as the target quantity.

[0023] The target pavement defect identification result is input into the prediction module. Based on the target pavement defect identification result, the prediction module obtains the time required for pavement defects with a severity of less than moderate to develop into pavement defects with a severity of moderate in the target pavement image. Then, the time required for pavement defects with a severity of less than moderate to develop into pavement defects with a severity of moderate in the target pavement image, as output by the prediction module, is obtained as the target time.

[0024] The target quantity and the target duration are input into the output module, and the output module obtains the pavement distress development prediction result based on the target duration and the target quantity, and then obtains the pavement distress development prediction result output by the output module.

[0025] According to a method for predicting the development of pavement defects provided by the present invention, the step of identifying pavement defects in the target pavement image and obtaining the type and geometric information of the pavement defects in the target pavement image as the target pavement defect identification result includes:

[0026] The target pavement image is input into the pavement defect recognition model to obtain the type and geometric information of the pavement defects in the target pavement image output by the pavement defect recognition model, which is used as the target pavement defect recognition result.

[0027] The pavement defect identification model is obtained by training on sample pavement images of the second sample road segment and the sample pavement defect identification results of the sample pavement images. The sample pavement defect identification results include sample pavement images labeled with pavement defects, pavement defect type information, and geometric information.

[0028] According to the method for predicting the development of pavement defects provided by the present invention, the time-series results of pavement defect identification in the sample are obtained based on the following steps:

[0029] According to the time sequence, the road surface image of each first sample road segment in the road surface time sequence image is sequentially input into the road surface defect identification model to obtain the road surface image of each first sample road segment labeled with road surface defects, the type information of the road surface defects and geometric information, which are sequentially output by the road surface defect identification model.

[0030] Based on the location of pavement defects in the pavement images of each first sample road segment, the location information of the pavement defects is marked in the pavement images of each first sample road segment that are labeled with pavement defects, the type information of the pavement defects, and geometric information. Then, the pavement images of each first sample road segment labeled with pavement defects, the type information of the pavement defects, the location information, and the geometric information are determined as the temporal results of the pavement defect identification of the sample.

[0031] According to a method for predicting pavement distress development provided by the present invention, after obtaining the pavement distress development prediction results of the road segment to be predicted output by the pavement distress development prediction model, the method further includes:

[0032] Based on the time-series variation data of the number of pavement defects with a severity of not less than moderate in the road segment to be predicted, the time required for the road segment to be predicted to reach the level of moderate defects is determined.

[0033] Wherein, if the road section to be predicted reaches the level of moderate damage, the target number must be no less than a preset value, and the target number is the number of road surface defects in the road section to be predicted that are no less severe than moderate.

[0034] The present invention also provides a pavement distress development prediction device, comprising:

[0035] The image acquisition module is used to acquire road surface images of the road segment to be predicted, which are then used as target road surface images.

[0036] The pavement defect identification module is used to identify pavement defects in the target pavement image, and to obtain the type information and geometric information of the pavement defects in the target pavement image as the target pavement defect identification result.

[0037] The pavement disease development prediction module is used to input the target pavement disease identification results into the pavement disease development prediction model and obtain the pavement disease development prediction results of the pavement section to be predicted output by the pavement disease development prediction model.

[0038] The pavement distress development prediction results include: time-series data on the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the pavement time-series images of the first sample road segment and the sample pavement distress identification time-series results of the pavement time-series images; the pavement time-series images include pavement images of the first sample road segment collected at different times and arranged in chronological order; the sample pavement distress identification time-series results include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distresses as well as the type, location, and geometric information of the pavement distresses.

[0039] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the pavement distress development prediction method as described above.

[0040] The present invention also provides a pavement distress development prediction system, comprising: an image sensor and the electronic device described above; the image sensor is electrically connected to the electronic device; the image sensor is disposed on a moving body;

[0041] The image sensor is used to acquire road surface images of the road segment to be predicted when the moving body moves to the road segment to be predicted, and to send the road surface images to the electronic device.

[0042] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the pavement distress development prediction method as described above.

[0043] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the pavement distress development prediction method as described above.

[0044] The method, apparatus, system, equipment, and medium for predicting road surface defects provided by this invention acquire a road surface image of the road segment to be predicted as the target road surface image, then identify road surface defects in the target road surface image, obtain the type and geometric information of the road surface defects in the target road surface image as the target road surface defect identification result, and then input the target road surface defect identification result into the road surface defect development prediction model to obtain the road surface defect development prediction result of the road segment to be predicted output by the road surface defect development prediction model. This can more accurately and efficiently predict the development trend of micro-level road surface defects, provide decision support for the timing of preventive maintenance, increase the service life of roads, and improve the operational efficiency of roads. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0046] Figure 1 This is a flowchart illustrating the method for predicting the development of road surface defects provided by this invention.

[0047] Figure 2 This is a flowchart illustrating the process of training a pavement defect identification model in the pavement defect development prediction method provided by the present invention.

[0048] Figure 3 This is a schematic diagram of the process for constructing a pavement distress development prediction model in the pavement distress development prediction method provided by the present invention;

[0049] Figure 4 This is a schematic diagram showing the relationship between the evaluation parameter values ​​corresponding to the pavement distress type and time in the pavement distress development prediction method provided by this invention.

[0050] Figure 5 This is a schematic diagram illustrating the relationship between the number of pavement defects of at least moderate severity in the road section to be predicted and time in the pavement defect development prediction method provided by this invention.

[0051] Figure 6 This is a schematic diagram of the pavement distress development prediction device provided by the present invention;

[0052] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0054] In the description of the invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0055] In the description of this application, the terms "first," "second," etc., are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, without limiting the number of objects; for example, a first object can be one or more. Furthermore, in the description of this application, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects have an "or" relationship.

[0056] Figure 1 This is a flowchart illustrating the pavement distress development prediction method provided by this invention. The following is a summary... Figure 1 This invention describes a method for predicting the development of pavement defects. For example... Figure 1 As shown, the method includes: step 101, acquiring the road surface image of the road segment to be predicted as the target road surface image.

[0057] It should be noted that the execution subject of this embodiment of the invention is a road surface distress development prediction device.

[0058] Specifically, the road segment to be predicted is the prediction object of the pavement distress development prediction method provided by this invention. Based on the pavement distress development prediction method provided by this invention, the development trend of micro-pavement distress on the road segment to be predicted can be predicted, and the pavement distress development prediction result of the road segment to be predicted can be obtained.

[0059] It is understood that, in the embodiments of the present invention, the road segment to be predicted can be determined according to actual needs.

[0060] Optionally, in this embodiment of the invention, the road surface of the road segment to be predicted can be an asphalt road surface.

[0061] It should be noted that the severity of pavement distress in this embodiment of the invention can be divided into mild, moderate, and severe. Based on the severity of the distress, pavement distress can be classified as microscopic pavement distress, moderate pavement distress, or severe pavement distress.

[0062] In this embodiment of the invention, the types of pavement defects and their correspondence with evaluation parameters and severity classification thresholds can be predefined based on prior knowledge and / or actual conditions. The severity classification thresholds may include a first threshold for classifying mild to moderate defects and a second threshold for classifying moderate to severe defects.

[0063] Based on the type of pavement distress and the corresponding relationship between evaluation parameters and severity thresholds, the severity of any pavement distress can be determined based on its type and geometric information. This allows us to determine whether the distress is a micro-level, moderate, or severe pavement distress.

[0064] The type information of pavement distress can be used to describe the type of pavement distress. In this embodiment of the invention, the types of pavement distress may include any multiple of the following: pavement cracking, pavement block cracking, pavement longitudinal cracking, pavement transverse cracking, pavement settlement, pavement rutting, pavement wave bulging, and pavement potholes.

[0065] In this embodiment of the invention, the geometric information of road surface defects may include at least one of the following: transverse length, longitudinal length, depth, and area.

[0066] The first threshold and evaluation parameters corresponding to different types of pavement distress are shown in Table 1.

[0067] Table 1. Correspondence between different types of pavement distress, evaluation parameters, and the first threshold.

[0068] Disease type K Evaluation parameters First threshold S <![CDATA[Pavement Cracking K1]]> Average crack width <![CDATA[S1=2mm]]> <![CDATA[Block crack K2]]> Average crack width <![CDATA[S2=2mm]]> <![CDATA[Vertical crack K3]]> Average crack width <![CDATA[S3=3mm]]> <![CDATA[Transverse crack K4]]> Average crack width <![CDATA[S4=3mm]]> <![CDATA[Pavement settlement K5]]> Settlement depth <![CDATA[S5=25mm]]> <![CDATA[Rutting K6]]> rut depth <![CDATA[S6=15mm]]> <![CDATA[Wave Heave K7]]> The difference in height between the crests and troughs <![CDATA[S7=25mm]]> <![CDATA[Pothole K8]]> Pot depth and / or pot area <![CDATA[S8 = 25 mm and / or 0.1 m 2 >

[0069] For a specific pavement distress type i, the evaluation parameters and first threshold corresponding to the distress can be determined based on Table 1. Then, by judging whether the evaluation parameters corresponding to the distress are greater than the corresponding first threshold, the severity of the distress can be determined as moderate. For example, in the case of pavement cracking, if the average crack width is less than 2mm based on the geometric information of the distress, the distress severity can be classified as mild, and the distress can be classified as a micro-disturbance. If the average crack width increases to 2mm based on the geometric information of the distress, the distress severity can be classified as moderate, and the distress can be classified as a moderate pavement distress.

[0070] In the embodiments of the present invention, it can be used N represents the number of pavement defects of type i with moderate severity at time t. t Let N represent the number of pavement defects with a severity level of moderate. t This can be expressed by the following formula:

[0071]

[0072] In this embodiment of the invention, a mobile three-dimensional image acquisition device can be used to acquire road surface images of the road section to be predicted, which can then be used as target road surface images.

[0073] Accordingly, in this embodiment of the invention, the target road surface image acquired by the mobile three-dimensional image acquisition device is a three-dimensional image.

[0074] Step 102: Perform pavement distress identification on the target pavement image to obtain the type and geometric information of pavement distress in the target pavement image, which will be used as the target pavement distress identification result.

[0075] Specifically, after acquiring the target road surface image, pavement defects can be identified through visual interpretation, deep learning, and other methods. This process obtains the type and geometric information of the pavement defects in the target pavement image, which serves as the target pavement defect identification result.

[0076] It is understandable that the number of pavement defects in the target pavement image can be one or more.

[0077] As an optional embodiment, pavement distress identification is performed on the target pavement image to obtain the type information and geometric information of the pavement distress in the target pavement image as the target pavement distress identification result, including: inputting the target pavement image into the pavement distress identification model, obtaining the type information and geometric information of the pavement distress in the target pavement image output by the pavement distress identification model as the target pavement distress identification result;

[0078] The pavement distress identification model is obtained by training on sample pavement images of the second sample road segment and the sample pavement distress identification results of the sample pavement images. The sample pavement distress identification results include sample pavement images labeled with pavement distress, pavement distress type information and geometric information.

[0079] Figure 2 This is a schematic diagram of the process for training the pavement defect identification model in the pavement defect development prediction method provided by this invention. For example... Figure 2 As shown, the specific steps for training the road surface defect identification model include: Step 21, using a mobile 3D image acquisition device to acquire sample road surface images of the second sample road segment.

[0080] Step 22: Identify pavement defects in the sample pavement images;

[0081] Step 23: Mark the pavement defects in the sample pavement image, as well as the type and geometric information of the pavement defects, and determine the marked sample pavement image as the sample pavement defect identification result;

[0082] Step 24: Using sample road surface images as samples and sample road surface defect identification results as sample labels, train the neural network model to obtain a road surface defect identification model.

[0083] Specifically, in this embodiment of the invention, multiple three-dimensional image acquisition devices can be installed on the mobile body, and these mobile three-dimensional image acquisition devices are arranged in a direction perpendicular to the vehicle body. During the parallel travel of the multiple mobile bodies along the second sample road segment, the three-dimensional image acquisition devices on each mobile body can simultaneously photograph the road surface of the second sample road segment, thereby acquiring road surface images of the second sample road segment as sample road surface images of the second sample road segment.

[0084] It should be noted that during the process of the aforementioned mobile vehicle traveling along the second sample road section, the aforementioned mobile vehicle 3D image acquisition devices can simultaneously take pictures of the road surface of the second sample road section, which can reduce distortion caused by camera distortion and thus improve the image quality of the sample road surface.

[0085] It should be noted that the three-dimensional image acquisition device in this embodiment of the invention is a mobile and portable three-dimensional image acquisition device.

[0086] Understandably, there can be multiple second sample road segments. The more second sample road segments there are, the higher the computational accuracy of the road surface defect identification model trained on it.

[0087] After obtaining the pavement distress identification model, the target pavement image can be input into the pavement distress identification model.

[0088] The pavement distress identification model can identify pavement distress in target pavement images, and obtain and output information on the type and geometry of pavement distress in the target pavement images.

[0089] After obtaining the type and geometric information of pavement defects in the target pavement image output by the pavement defect recognition model, the type and geometric information of pavement defects in the target pavement image can be determined as the target pavement defect recognition result.

[0090] The pavement defect identification model in this embodiment of the invention is obtained after training based on sample pavement images and sample pavement defect identification results. The sample pavement images include sample pavement images of a second sample road segment, and the sample pavement defect identification results include the type information and geometric information of pavement defects on the pavement of the second sample road segment. This embodiment of the invention improves the identification efficiency and accuracy of pavement defect identification on target pavement images by inputting the target pavement image into the pavement defect identification model and obtaining the type information and geometric information of pavement defects in the target pavement image output by the pavement defect identification model as the target pavement defect identification result. It can reduce the cost required for pavement defect identification on target pavement images and obtain the type information and geometric information of pavement defects in target pavement images more accurately and efficiently.

[0091] Step 103: Input the target pavement distress identification results into the pavement distress development prediction model to obtain the pavement distress development prediction results of the road segment to be predicted output by the pavement distress development prediction model.

[0092] The pavement distress development prediction results include: time-series data on the number of pavement distresses of severity not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the pavement time-series images of the first sample road segment and the sample pavement distress identification time-series results of the pavement time-series images; the pavement time-series images include pavement images of the first sample road segment collected at different times and arranged in chronological order; the sample pavement distress identification time-series results include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distresses and information on their type, location, and geometry.

[0093] Figure 3 This is a schematic diagram illustrating the process of constructing a pavement distress development prediction model in the pavement distress development prediction method provided by this invention. For example... Figure 3 As shown, the specific steps for constructing a pavement distress development prediction model include: Step 31, using a mobile 3D image acquisition device to acquire pavement images of the first sample road section multiple times to obtain the pavement time-series images of the first sample road section.

[0094] Step 32: Obtain the pavement image of each first sample road segment, which is arranged in chronological order and labeled with pavement defects, the type information of the aforementioned pavement defects, location information and geometric information, as the sample pavement defect identification chronological result of the aforementioned pavement chronological image.

[0095] Step 33: Based on the time series results of pavement distress identification in the sample, calculate the time required for each type of pavement distress from its appearance until its severity changes to moderate.

[0096] Step 34: Based on the time-series results of pavement distress identification in the sample, calculate the time required for each type of pavement distress from its appearance until its severity changes to moderate.

[0097] Specifically, in this embodiment of the invention, multiple mobile 3D image acquisition devices can be installed on a mobile vehicle, and these mobile 3D image acquisition devices are arranged in a direction perpendicular to the vehicle body. When acquiring road surface images of the first sample road segment, the multiple mobile vehicles can travel parallel to the first sample road segment. During this parallel travel, the 3D image acquisition devices on each mobile vehicle can simultaneously photograph the road surface of the first sample road segment, thereby acquiring the road surface image of the first sample road segment and recording the acquisition time of the road surface image of the first sample road segment.

[0098] It should be noted that, in this embodiment of the invention, the time interval between the current acquisition of the road surface image of the first sample road segment and the previous acquisition of the road surface image of the first sample road segment is a preset time.

[0099] It is understandable that the road surface time-series images of the first sample road segment include multiple road surface images of the first sample road segment collected at different times and arranged in chronological order.

[0100] After acquiring the time-series images of the first sample road segment, in this embodiment of the invention, based on the time-series images of the first sample road segment, through visual interpretation, deep learning, and other methods, the road surface images of each first sample road segment, which are arranged in time sequence and labeled with road surface defects, the type information, location information, and geometric information of the aforementioned road surface defects, can be acquired as the time-series result of the sample road surface defect identification of the aforementioned time-series images.

[0101] As an optional embodiment, the temporal results of sample pavement defects identification are obtained based on the following steps: according to the temporal order, the pavement images of each first sample road segment in the pavement temporal image are sequentially input into the pavement defect identification model, and the pavement images of each first sample road segment labeled with pavement defects, pavement defect type information and geometric information are sequentially output by the pavement defect identification model.

[0102] Based on the location of pavement defects in the pavement image of each first sample road segment, the location information of pavement defects is marked in the pavement image of each first sample road segment labeled with pavement defects, pavement defect type information, and geometric information. Then, the pavement image of each first sample road segment labeled with pavement defects, pavement defect type information, location information, and geometric information is determined as the sample pavement defect identification time sequence result.

[0103] The embodiments of the present invention obtain the time-series results of sample road surface defects by using a road surface defect identification model. This can improve the identification efficiency and accuracy of road surface defects in the time-series images of the first sample road segment, reduce the cost required for road surface defects in the time-series images of the first sample road segment, and obtain the time-series results of sample road surface defects more accurately and efficiently.

[0104] As an optional embodiment, the pavement distress development prediction model includes: an identification module, a prediction module, and an output module; the prediction module is constructed based on the following steps: based on the time series results of sample pavement distress identification, the time required for each type of pavement distress to change from its appearance to its severity becoming moderate is obtained;

[0105] Among them, the correspondence between the types of pavement defects and the evaluation parameters and severity classification thresholds is predefined, and the severity classification thresholds include a first threshold used to classify mild and moderate defects.

[0106] Specifically, after obtaining the time-series results of sample pavement defects identification, the time required for each type of pavement defect to change from its appearance to its severity becoming moderate can be calculated based on these results using deep learning technology, manual statistics, and other methods.

[0107] As an optional embodiment, based on the temporal results of sample pavement distress identification, the time required for each type of pavement distress to change from its appearance to its severity becoming moderate is obtained, including: based on the location information of pavement distress in the pavement images of each first sample road segment, performing entity alignment on the pavement distress in the pavement images of each first sample road segment, and associating the same pavement distress in the pavement images of each first sample road segment based on the entity alignment results.

[0108] Specifically, entity alignment (EA), also known as entity matching or entity resolution, is one of the most fundamental and critical technologies in knowledge fusion. The goal of entity alignment is to match and correspond identical entities in different images.

[0109] In this embodiment of the invention, road surface defects can be treated as entities. Entity alignment is performed on the road surface defects in the road surface images of each first sample road segment. For example, road surface defect 'a' in the road surface image of the first sample road segment acquired at acquisition time A and road surface defect 'b' in the road surface image of the first sample road segment acquired at acquisition time B are the same road surface defect on the road surface of the first sample road segment. By performing entity alignment on the road surface defects in the road surface images of each first sample road segment, it can be determined that road surface defect 'a' in the road surface image of the first sample road segment acquired at acquisition time A and road surface defect 'b' in the road surface image of the first sample road segment acquired at acquisition time B are the same road surface defect.

[0110] After obtaining the temporal results of pavement defect identification in the sample road surface, in this embodiment of the invention, based on the location information of pavement defects in the pavement images of each first sample road segment, entity alignment of the pavement defects in the pavement images of each first sample road segment can be performed through numerical comparison or deep learning techniques. Then, based on the entity alignment results, the same pavement defects can be managed in the pavement images of each first sample road segment. For example, if it is determined that pavement defect 'a' in the pavement image of the first sample road segment acquired at acquisition time A is the same pavement defect 'b' in the pavement image of the first sample road segment acquired at acquisition time B, then pavement defect 'a' in the pavement image of the first sample road segment acquired at acquisition time A can be associated with pavement defect 'b' in the pavement image of the first sample road segment acquired at acquisition time B.

[0111] For each identical pavement defect, based on the geometric and type information of each identical pavement defect in the pavement images of different first sample road segments, the severity of each identical pavement defect is determined to be moderate in the pavement images of the first sample road segments.

[0112] Specifically, in the embodiments of the present invention, i represents any identical pavement distress.

[0113] For pavement distress i, based on the type information of pavement distress i, the evaluation parameters corresponding to pavement distress i and the first threshold used to classify it as mild or moderate can be determined.

[0114] If, based on the geometric information of pavement distress i in the pavement image of a certain first sample road segment, it is determined that the evaluation parameter value corresponding to pavement distress i exceeds the first threshold corresponding to pavement distress i, then the severity of pavement distress i in the pavement image of the aforementioned first sample road segment can be determined to be moderate.

[0115] Based on the temporal order of the pavement images of each first sample road segment, the pavement images of the first sample road segment whose severity of pavement distress i first changes to moderate can be determined.

[0116] Based on the acquisition time of the pavement image of the first sample road segment where each identical pavement defect first appears, and the acquisition time of the pavement image of the first sample road segment where the severity of each identical pavement defect first changes to moderate, the time required for each identical pavement defect to change from appearance to moderate severity is obtained.

[0117] Specifically, in this embodiment of the invention, the time interval between the acquisition time of the pavement image of the first sample road segment where pavement distress i first appears and the acquisition time of the pavement image of the first sample road segment where the severity of pavement distress i first changes to moderate can be calculated.

[0118] After obtaining the above interval duration, the above interval duration can be directly determined as the time required for pavement distress i to change from severity to moderate from the time it appears. Alternatively, the sum of the above interval duration and the preset time delay can be determined as the time required for pavement distress i to change from severity to moderate from the time it appears.

[0119] Based on the time required for each identical pavement distress to change from severity to moderate from its appearance, the time required for each type of pavement distress to change from severity to moderate from its appearance is obtained.

[0120] Specifically, after obtaining the time required for each identical pavement distress to change from its appearance to a moderate severity, the same pavement distress can be classified based on its type. Then, the average time required for each identical pavement distress within the same type to change from its appearance to a moderate severity can be calculated numerically, which serves as the average time required for the pavement distress of the aforementioned type to change from its appearance to a moderate severity.

[0121] A prediction module is constructed based on the time required for each type of pavement distress to change from severe to moderate.

[0122] Specifically, in this embodiment of the invention, j represents any type of pavement distress.

[0123] For pavement distress type j, after obtaining the time required for the severity of pavement distress of type j to change from its appearance to moderate, the evaluation parameter values ​​corresponding to type j can be linearly fitted based on the time required for the severity of pavement distress of type j to change from its appearance to moderate and the first threshold corresponding to type j, so as to obtain the time series change data of the evaluation parameter values ​​corresponding to pavement distress type j as they change over time.

[0124] Figure 4 This is a schematic diagram illustrating the relationship between the evaluation parameter values ​​corresponding to pavement distress types and time in the pavement distress development prediction method provided by this invention. The relationship between the evaluation parameter values ​​corresponding to pavement distress types and time is as follows: Figure 4 As shown.

[0125] For example, if the type of pavement distress is pavement cracking, and the time required for the pavement distress to change from severe to moderate is determined to be t, then the average crack width (the evaluation parameter corresponding to pavement cracking) value can be linearly fitted based on the time t and 2 mm (the first threshold corresponding to pavement cracking) to obtain the time series change data of the average crack width value corresponding to pavement cracking as it changes over time.

[0126] Based on the time-series data showing the changes in the evaluation parameter values ​​corresponding to each type over time, the prediction module in the pavement distress development prediction model can be constructed.

[0127] As an optional embodiment, the target pavement distress identification result is input into the pavement distress development prediction model to obtain the pavement distress development prediction result of the road segment to be predicted output by the pavement distress development prediction model. This includes: inputting the target pavement distress identification result into the identification module, whereby the identification module determines the number of pavement distresses with a severity of not less than moderate in the target pavement image based on the target pavement distress identification result, and then obtaining the number of pavement distresses with a severity of not less than moderate in the target pavement image output by the identification module as the target number; inputting the target pavement distress identification result into the prediction module, whereby the prediction module obtains the time required for pavement distresses with a severity of less than moderate to develop into pavement distresses with a severity of moderate based on the target pavement distress identification result, and then obtaining the time required for pavement distresses with a severity of less than moderate to develop into pavement distresses with a severity of moderate in the target pavement image output by the prediction module as the target time.

[0128] Specifically, after the target pavement distress identification results are input into the identification module of the pavement distress development prediction model, the identification module can determine and output the number of pavement distresses with a severity of not less than moderate in the target pavement image based on the geometric and type information of the pavement distresses in the target pavement image, according to the correspondence between the type of pavement distress and the evaluation parameters and severity classification thresholds.

[0129] After obtaining the number of pavement defects with a severity level of not less than moderate in the target pavement image output by the recognition module, the above can be determined as the target number.

[0130] After the target pavement distress identification results are input into the prediction module of the pavement distress development prediction model, the prediction module can obtain and output the time required for pavement distress with a severity of less than moderate to develop into pavement distress with a severity of moderate based on the geometric and type information of pavement distress in the target pavement image and the time-series change data of the evaluation parameter values ​​corresponding to each type over time.

[0131] After obtaining the time required for pavement distress of less than moderate severity to develop into pavement distress of moderate severity in the target pavement image output by the prediction module, the above time can be determined as the target time.

[0132] Input the target quantity and target duration into the output module, and obtain the pavement distress development prediction results output by the output module.

[0133] Specifically, after obtaining the target quantity and target duration, the target quantity and target duration can be input into the output module.

[0134] The output module can use time as the independent variable, the number of pavement defects of severity no less than moderate in the road segment to be predicted as the dependent variable, the time when the target pavement image is collected as the time when the independent variable is 0, and the above target number as the starting value of the dependent variable when the independent variable is 0. Based on the time required for pavement defects of less than moderate severity in the target pavement image to develop into pavement defects of moderate severity, when the time reaches the time required for a pavement defect of less than moderate severity in the target pavement image to develop into pavement defects of moderate severity, the dependent variable is incremented by 1, thus obtaining the time-series change data of the number of pavement defects of severity no less than moderate in the road segment to be predicted.

[0135] It should be noted that the output module can also use time as the independent variable, the number of each type of pavement distress with a severity of not less than moderate in the road segment to be predicted as the dependent variable, the time when the target pavement image is collected as the time when the independent variable is 0, and the number of each type of pavement distress in the above target quantity as the starting value of the dependent variable when the independent variable is 0. Based on the time required for pavement distress with a severity of less than moderate in the target pavement image to develop into pavement distress with a severity of moderate, when the time reaches the time required for pavement distress with a severity of less than moderate in the target pavement image to develop into pavement distress with a severity of moderate, the number of each type of pavement distress is incremented by 1. This can obtain the time-series change data of the number of each type of pavement distress with a severity of not less than moderate in the road segment to be predicted.

[0136] Figure 5 This is a schematic diagram illustrating the relationship between the number of pavement defects of at least moderate severity in the road segment to be predicted and time, as provided in the pavement defect development prediction method of this invention. The relationship between the number of pavement defects of at least moderate severity in the road segment to be predicted and time is as follows: Figure 5 As shown.

[0137] As an optional embodiment, after obtaining the pavement distress development prediction results of the road segment to be predicted output by the pavement distress development prediction model, the method further includes: determining the time required for the road segment to be predicted to reach the level of moderate distress based on the time-series change data of the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted.

[0138] For a road section to be predicted to reach a moderate level of road damage, the target number must be no less than a preset value. The target number is the number of road surface defects in the predicted road section that are no less severe than moderate.

[0139] Specifically, based on the time-series change data of the number of pavement defects of severity no less than moderate in the road segment to be predicted, the time required for the number of pavement defects of severity no less than moderate in the road segment to be predicted to be no less than a preset value can be obtained through mathematical statistics. Then, the above time can be determined as the time required for the road segment to be predicted to reach the level of moderate defects.

[0140] It is understandable that the time required for the predicted road section to reach moderate damage level can be 0 or greater than 0.

[0141] This invention, through obtaining a pavement image of the road segment to be predicted as the target pavement image, performs pavement distress identification on the target pavement image, obtains the type and geometric information of the pavement distress in the target pavement image as the target pavement distress identification result, and then inputs the target pavement distress identification result into a pavement distress development prediction model to obtain the pavement distress development prediction result of the road segment to be predicted output by the pavement distress development prediction model. This can more accurately and efficiently predict the development trend of micro-pavement distress, provide decision support for the timing of preventive maintenance, increase the service life of roads, and improve the operational efficiency of roads.

[0142] Figure 6 This is a schematic diagram of the pavement distress development prediction device provided by the present invention. The following is in conjunction with... Figure 6 The pavement distress development prediction device provided by this invention is described below. The pavement distress development prediction device described below can be referred to in correspondence with the pavement distress development prediction method provided by this invention described above. For example... Figure 6 As shown, there is an image acquisition module 601, a disease identification module 602, and a disease development prediction module 603.

[0143] The image acquisition module 601 is used to acquire road surface images of the road segment to be predicted, as the target road surface image;

[0144] The defect identification module 602 is used to identify road defects in the target road image, and to obtain the type information and geometric information of the road defects in the target road image as the target road defect identification result.

[0145] The pavement disease development prediction module 603 is used to input the target pavement disease identification result into the pavement disease development prediction model and obtain the pavement disease development prediction result of the pavement section to be predicted output by the pavement disease development prediction model.

[0146] The pavement distress development prediction results include: time-series data on the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the pavement time-series images of the first sample road segment and the sample pavement distress identification time-series results of the pavement time-series images; the pavement time-series images include pavement images of the first sample road segment collected at different times and arranged in chronological order; the sample pavement distress identification time-series results include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distresses as well as the type, location, and geometric information of the pavement distresses.

[0147] Specifically, the image acquisition module 601, the disease identification module 602, and the disease development prediction module 603 are electrically connected.

[0148] The pavement distress development prediction device in this embodiment of the invention acquires a pavement image of the road segment to be predicted as the target pavement image, then identifies pavement distress in the target pavement image, obtains the type and geometric information of the pavement distress in the target pavement image as the target pavement distress identification result, and then inputs the target pavement distress identification result into the pavement distress development prediction model to obtain the pavement distress distress development prediction result of the road segment to be predicted output by the pavement distress distress development prediction model. It can more accurately and efficiently predict the development trend of micro pavement distress, provide decision support for the timing of preventive maintenance, increase the service life of the road, and improve the operating efficiency of the road.

[0149] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7As shown, the electronic device may include: a processor 710, a communication interface 720, a memory 730, and a communication bus 740, wherein the processor 710, the communication interface 720, and the memory 730 communicate with each other through the communication bus 740. The processor 710 can call logic instructions in the memory 730 to execute a pavement distress development prediction method. This method includes: acquiring pavement images of the road segment to be predicted, as target pavement images; performing pavement distress identification on the target pavement images to obtain type and geometric information of the pavement distress in the target pavement images, as target pavement distress identification results; inputting the target pavement distress identification results into a pavement distress development prediction model to obtain the pavement distress distress development prediction results of the road segment to be predicted output by the pavement distress distress development prediction model; wherein the pavement distress distress development prediction results include: time-series variation data of the number of pavement distresses with a severity not less than moderate in the road segment to be predicted; the pavement distress distress development prediction model is constructed based on the pavement time-series images of a first sample road segment and the sample pavement distress identification time-series results of the pavement time-series images; the pavement time-series images include pavement images of multiple first sample road segments acquired at different times and arranged in chronological order; the sample pavement distress identification time-series results include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distress, pavement distress type information, location information, and geometric information.

[0150] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0151] Based on the above embodiments, a road surface defect development prediction system includes: an image sensor and an electronic device as described above; the image sensor is electrically connected to the electronic device; the image sensor is disposed on a moving body; the image sensor is used to acquire a road surface image of the road segment to be predicted when the moving body moves to the road segment to be predicted, and to send the road surface image of the road segment to be predicted to the electronic device.

[0152] Specifically, the pavement distress development prediction system in this embodiment of the invention includes an image sensor and an electronic device. The image sensor is mounted on a moving body and, when the moving body moves to the road segment to be predicted, acquires a pavement image of the road segment to be predicted and sends the pavement image of the road segment to be predicted to the electronic device. The electronic device then uses the pavement distress development prediction method provided by this invention to obtain the pavement distress development prediction result of the road segment to be predicted based on the pavement image of the road segment to be predicted.

[0153] It should be noted that the specific steps of the electronic device executing the road surface distress development prediction method provided by the present invention can be found in the above embodiments, and will not be repeated in the embodiments of the present invention.

[0154] The pavement distress development prediction system in this invention can be installed on ordinary passenger cars, thus eliminating the high cost of road inspection vehicles and enabling independent inspection work. This significantly improves the frequency and accuracy of road distress identification while maintaining the economic efficiency of data collection. Therefore, predicting the development of micro-diseases in asphalt pavements helps address the insufficient predictive capabilities of existing technologies, providing decision support for preventative maintenance timing, improving the operational efficiency of target roads, and extending their service life.

[0155] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the pavement distress development prediction method provided by the above methods. The method includes: acquiring a pavement image of the road segment to be predicted as a target pavement image; identifying pavement distress in the target pavement image to obtain the type information and geometric information of the pavement distress in the target pavement image as the target pavement distress identification result; and inputting the target pavement distress identification result into a pavement distress development prediction model to obtain a pavement distress development prediction. The model outputs the pavement distress development prediction results for the road segment to be predicted; among which, the pavement distress development prediction results include: the temporal variation data of the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the pavement temporal images of the first sample road segment and the sample pavement distress identification temporal results of the pavement temporal images; the pavement temporal images include pavement images of the first sample road segment with different acquisition times and arranged in temporal order; the sample pavement distress identification temporal results include pavement images of each first sample road segment arranged in temporal order and labeled with pavement distresses and pavement distress type information, location information and geometric information.

[0156] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program performs the pavement distress development prediction method provided by the methods described above. This method includes: acquiring a pavement image of a road segment to be predicted, as a target pavement image; performing pavement distress identification on the target pavement image, obtaining type information and geometric information of the pavement distress in the target pavement image, as a target pavement distress identification result; inputting the target pavement distress identification result into a pavement distress development prediction model, and obtaining the pavement distress data of the road segment to be predicted output by the pavement distress development prediction model. The pavement distress development prediction results include: time-series data on the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the pavement time-series images of the first sample road segment and the sample pavement distress identification time-series results of the pavement time-series images; the pavement time-series images include pavement images of the first sample road segment collected at different times and arranged in chronological order; the sample pavement distress identification time-series results include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distresses and information on their type, location, and geometry.

[0157] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0158] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0159] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention 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; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the development of pavement defects, characterized in that, include: Obtain road surface images of the road segment to be predicted, and use them as target road surface images; The target pavement image is subjected to pavement distress identification to obtain the type and geometric information of the pavement distress in the target pavement image, which is used as the target pavement distress identification result. The target pavement distress identification results are input into the pavement distress development prediction model to obtain the pavement distress development prediction results of the road segment to be predicted output by the pavement distress development prediction model. The pavement distress development prediction results include: time-series data on the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the time-series images of the first sample road segment and the time-series results of sample pavement distress identification from the time-series images; the time-series images of the pavement include pavement images of multiple first sample road segments collected at different times and arranged in chronological order; the time-series results of sample pavement distress identification include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distresses, as well as the type, location, and geometric information of the pavement distresses. The pavement distress development prediction model includes: an identification module, a prediction module, and an output module; The prediction module is constructed based on the following steps: Based on the time-series results of pavement distress identification in the sample, the time required for each type of pavement distress to change from its appearance to its severity becoming moderate is obtained. The prediction module is constructed based on the time required for each type of pavement distress to change from its appearance to a moderate severity. Among them, the correspondence between the types of pavement defects and the evaluation parameters and severity classification thresholds is predefined, and the severity classification thresholds include a first threshold for classifying mild and moderate defects. The step of obtaining the time required for each type of pavement distress to change from its appearance to moderate severity based on the time-series results of the pavement distress identification sample includes: Based on the location information of pavement defects in the pavement images of each first sample road segment, the pavement defects in the pavement images of each first sample road segment are aligned with entities, and the same pavement defects are associated in the pavement images of each first sample road segment based on the entity alignment results. For each identical pavement defect, based on the geometric and type information of each identical pavement defect in the pavement images of different first sample road segments, the pavement images of the first sample road segments whose severity first becomes moderate are determined. Based on the acquisition time of the first sample road segment of each identical pavement defect when it first appears, and the acquisition time of the first sample road segment of each identical pavement defect when its severity first changes to moderate, the time required for each identical pavement defect to change from its appearance to its severity becoming moderate is obtained. Based on the time required for each identical pavement distress to change from severity to moderate from its appearance, the time required for each type of pavement distress to change from severity to moderate from its appearance is obtained.

2. The method for predicting the development of pavement defects according to claim 1, characterized in that, The step of inputting the target pavement distress identification result into the pavement distress development prediction model and obtaining the pavement distress development prediction result of the road segment to be predicted output by the pavement distress development prediction model includes: The target pavement defect identification results are input into the identification module. Based on these results, the identification module determines the number of pavement defects in the target pavement image with a severity level of not less than moderate. The number of pavement defects with a severity level of not less than moderate in the target pavement image, output by the identification module, is then used as the target quantity. The target pavement defect identification result is input into the prediction module. Based on the target pavement defect identification result, the prediction module obtains the time required for pavement defects with a severity of less than moderate to develop into pavement defects with a severity of moderate in the target pavement image. Then, the time required for pavement defects with a severity of less than moderate to develop into pavement defects with a severity of moderate in the target pavement image, as output by the prediction module, is obtained as the target time. The target quantity and the target duration are input into the output module, and the output module obtains the pavement distress development prediction result based on the target duration and the target quantity, and then obtains the pavement distress development prediction result output by the output module.

3. The method for predicting the development of pavement defects according to claim 1, characterized in that, The step of identifying pavement defects in the target pavement image and obtaining the type and geometric information of the pavement defects in the target pavement image as the target pavement defect identification result includes: The target pavement image is input into the pavement defect recognition model to obtain the type and geometric information of the pavement defects in the target pavement image output by the pavement defect recognition model, which is used as the target pavement defect recognition result. The pavement defect identification model is obtained by training on sample pavement images of the second sample road segment and the sample pavement defect identification results of the sample pavement images. The sample pavement defect identification results include sample pavement images labeled with pavement defects, pavement defect type information, and geometric information.

4. The method for predicting the development of pavement defects according to claim 3, characterized in that, The time-series results of the sample pavement defect identification were obtained based on the following steps: According to the time sequence, the road surface image of each first sample road segment in the road surface time sequence image is sequentially input into the road surface defect identification model to obtain the road surface image of each first sample road segment labeled with road surface defects, the type information of the road surface defects and geometric information, which are sequentially output by the road surface defect identification model. Based on the location of pavement defects in the pavement images of each first sample road segment, the location information of the pavement defects is marked in the pavement images of each first sample road segment that are labeled with pavement defects, the type information of the pavement defects, and geometric information. Then, the pavement images of each first sample road segment labeled with pavement defects, the type information of the pavement defects, the location information, and the geometric information are determined as the temporal results of the pavement defect identification of the sample.

5. The method for predicting the development of pavement defects according to any one of claims 1 to 4, characterized in that, After obtaining the pavement distress development prediction results of the road segment to be predicted output by the pavement distress development prediction model, the method further includes: Based on the time-series variation data of the number of pavement defects with a severity of not less than moderate in the road segment to be predicted, the time required for the road segment to be predicted to reach the level of moderate defects is determined. Wherein, if the road section to be predicted reaches the level of moderate damage, the target number must be no less than a preset value, and the target number is the number of road surface defects in the road section to be predicted that are no less severe than moderate.

6. A pavement distress development prediction device, characterized in that, include: The image acquisition module is used to acquire road surface images of the road segment to be predicted, which are then used as target road surface images. The pavement defect identification module is used to identify pavement defects in the target pavement image, and to obtain the type information and geometric information of the pavement defects in the target pavement image as the target pavement defect identification result. The pavement disease development prediction module is used to input the target pavement disease identification results into the pavement disease development prediction model and obtain the pavement disease development prediction results of the pavement section to be predicted output by the pavement disease development prediction model. The pavement distress development prediction results include: time-series data on the number of pavement distresses with a severity of not less than moderate in the road segment to be predicted; the pavement distress development prediction model is constructed based on the time-series images of the first sample road segment and the time-series results of sample pavement distress identification from the time-series images; the time-series images of the pavement include pavement images of multiple first sample road segments collected at different times and arranged in chronological order; the time-series results of sample pavement distress identification include pavement images of each first sample road segment arranged in chronological order and labeled with pavement distresses, as well as the type, location, and geometric information of the pavement distresses. The pavement distress development prediction model includes: an identification module, a prediction module, and an output module; The prediction module is constructed based on the following steps: Based on the time-series results of pavement distress identification in the sample, the time required for each type of pavement distress to change from its appearance to its severity becoming moderate is obtained. The prediction module is constructed based on the time required for each type of pavement distress to change from its appearance to a moderate severity. Among them, the correspondence between the types of pavement defects and the evaluation parameters and severity classification thresholds is predefined, and the severity classification thresholds include a first threshold for classifying mild and moderate defects. The step of obtaining the time required for each type of pavement distress to change from its appearance to moderate severity based on the time-series results of the pavement distress identification sample includes: Based on the location information of pavement defects in the pavement images of each first sample road segment, the pavement defects in the pavement images of each first sample road segment are aligned with entities, and the same pavement defects are associated in the pavement images of each first sample road segment based on the entity alignment results. For each identical pavement defect, based on the geometric and type information of each identical pavement defect in the pavement images of different first sample road segments, the pavement images of the first sample road segments whose severity first becomes moderate are determined. Based on the acquisition time of the first sample road segment of each identical pavement defect when it first appears, and the acquisition time of the first sample road segment of each identical pavement defect when its severity first changes to moderate, the time required for each identical pavement defect to change from its appearance to its severity becoming moderate is obtained. Based on the time required for each identical pavement distress to change from severity to moderate from its appearance, the time required for each type of pavement distress to change from severity to moderate from its appearance is obtained.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the pavement distress development prediction method as described in any one of claims 1 to 5.

8. A pavement distress development prediction system, characterized in that, include: An image sensor and the electronic device as described in claim 7; the image sensor is electrically connected to the electronic device; the image sensor is disposed on a moving body; The image sensor is used to acquire road surface images of the road segment to be predicted when the moving body moves to the road segment to be predicted, and to send the road surface images to the electronic device.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the pavement distress development prediction method as described in any one of claims 1 to 5.

Images