Pavement distress inspection data analysis system based on internet of things and artificial intelligence

Abstract

Provided is a pavement distress inspection data analysis system based on Internet of Things and artificial intelligence. The pavement distress inspection data analysis system based on Internet of Things and artificial intelligence comprises: a central server, an inspection planning module, and a distress recognition module. By retrieving a set of distress types from the central server, a plurality of inspection paths including a main road are automatically generated; a comprehensive analysis of traffic flow, pavement distress, etc. on the paths is conducted to obtain priority indexes, efficiency analysis is performed on inspection devices so as to obtain inspection values, and accordingly, allocation of path inspection tasks is performed, thereby forming a dynamic path inspection plan. By performing distress recognition on collected pavement images, distress types and distress levels can be accurately extracted from the pavement images, thereby providing a quantitative distress assessment; and then deduplication-based recognition is performed on pavement images in which pavement distress is present, such that the similarity between a target image and comparison images can be accurately determined, thereby avoiding repeated inspection and improving the efficiency and accuracy of distress inspection.

Description

Road Defect Inspection Data Analysis System Based on Internet of Things and Artificial Intelligence Technical Field

[0001] This invention relates to the field of road defect identification technology, specifically a road defect inspection data analysis system based on the Internet of Things and artificial intelligence. Background Technology

[0002] Road defects refer to various forms of damage to road surfaces that occur over long-term use due to external factors such as vehicle load, climate change, construction quality, and material aging. These defects affect the service life of the road, driving safety, and ride comfort, and require timely detection and repair.

[0003] Traditional traffic and municipal road maintenance and inspection work involves staff driving vehicles to inspect road surface defects and roadside assets. When defects are found, the vehicle is pulled over, and the inspection task is completed by measuring, taking photos, uploading, and filling out an inspection form. The data is then compiled, statistically analyzed, and subsequent maintenance work is arranged. Obviously, this traditional inspection method is inefficient and unreliable, making it difficult to cope with the current situation of rapid traffic development. Summary of the Invention

[0004] The purpose of this invention is to provide a road defect inspection data analysis system based on the Internet of Things and artificial intelligence to solve the problems mentioned in the background art.

[0005] The objective of this invention can be achieved through the following technical solution: a road defect inspection data analysis system based on the Internet of Things and artificial intelligence, comprising: a central server, an inspection planning module, and a defect identification module;

[0006] The central server interacts with inspection equipment via a communication network. Various sensors on the inspection equipment patrol along corresponding paths to collect real-time road images and their corresponding location information, uploading this data to the central server for storage. Furthermore, the central server maintains several sets of road defect types. Each set stores images of the corresponding road defect type, along with their location information, three key features, defect parameters, and defect severity values. The central server also communicates with the mobile devices of road maintenance personnel, sending information about road defects, their location, and defect type to their mobile devices. When a maintenance is completed, the defect is removed from the corresponding defect type set.

[0007] The inspection planning module plans inspection paths based on various types of defects to guide inspection equipment in inspection and road image acquisition.

[0008] The road defect identification module extracts features from road images and identifies defects to determine whether road defects exist. Then, it performs deduplication on road images containing defects. The specific deduplication process is as follows:

[0009] Step 1: Select any one of the road images with road defects as the target image. Use each defect image in the corresponding defect type set as the comparison image. Retrieve the geometric contours of the target image and the comparison image and find the center point of the geometric contour. Align the center points of the geometric contours of the target image and the comparison image and rotate them 360° until the overlap area is the largest. Calculate the intersection area and union area under this overlap pattern and denote them as Q1 and Q2, respectively.

[0010] Step 2: Match the disease parameters corresponding to the target image and the comparison image one by one to obtain several parameter pairs, and calculate the difference between the parameters of each parameter pair to obtain the parameter difference value of each parameter pair; calculate the mean of the parameter difference between the target image and the comparison image by averaging the parameter differences of each parameter pair, and denote it as C;

[0011] Retrieve the positional information corresponding to the target image and the comparison image, calculate the spatial distance between them, and denot it as H;

[0012] Step 3: Retrieve the disease severity values ​​corresponding to the target image and the comparison image, and record them as S1 and S2 respectively;

[0013] Step 4: Normalize the intersection area Q1, union area Q2, mean parameter difference C, spatial spacing H, the severity value S1 of the target image, and the severity value S2 of the compared image, and take their values. Then, calculate the repetition index SQ using a formula. The specific formula is as follows:

[0014] Where η1, η2, η3, and η4 are the set weight constants; thus, the repetition index between the target image and each image in the corresponding disease type set can be obtained. The repetition index is compared and analyzed with the set repetition threshold. When the repetition index is greater than the set repetition threshold, the image is counted as a repetition. When the cumulative number of repetitions is greater than or equal to one, it means that this road image has been collected in the previous inspection, and it is deleted. When the cumulative number of repetitions is zero, the road image with road disease is recorded as a disease image, and it, along with its corresponding location information, three feature elements, disease parameters, and disease severity value, is stored in the corresponding disease type set of the central server.

[0015] Preferably, the specific process for disease identification is as follows:

[0016] The acquired road images are preprocessed to parse and associate the image location information, so that each road image is accurately matched with its corresponding location information. Geometric features, texture features and color features are extracted from the road images to obtain the three feature elements of the road images, which include geometric features, texture features and color features.

[0017] Based on the labeled sample dataset, a convolutional neural network is trained to identify different types of road defects. The three feature elements are input into the trained convolutional neural network algorithm model to output the defect type and its corresponding defect parameters. A weighting coefficient, denoted as Pi, is assigned to each defect parameter for each defect type. The defect parameters for each defect type are denoted as Wi. The defect severity value S is calculated by weighting the parameters according to the corresponding weighting coefficients. The specific calculation formula is as follows:

[0018] Where i = 1, 2, 3...I, I is a positive integer, I is the total number of disease parameters in a certain disease type, and i is any one of the disease parameters; thus, disease identification can be performed on each road image to filter out road images with road defects, and output the disease type and disease severity value corresponding to each road image with road defects.

[0019] Preferably, the specific process of planning inspection routes based on various types of disease sets is as follows:

[0020] Step 1: Retrieve the disease images of each road disease type stored in each disease type set, as well as the location information and disease severity value of each disease image. Mark the disease corresponding to each disease image at the corresponding location on the city map. Generate several paths according to the path algorithm, so that all road diseases are distributed to each path.

[0021] Step 2: Analyze the saturation level of traffic flow on the route to obtain the route saturation value, denoted as Bα;

[0022] Step 3: Retrieve the number of road defects along the route, as well as the defect type and severity value for each defect; set a base value for each defect type; match each road defect along the route with all the set defect types to obtain the corresponding base value; normalize the base value and severity value and take their numerical values; perform weighted calculation on the numerical values ​​to obtain the defect risk value for each road defect; sum the defect risk values ​​of each road defect along the route to obtain the total risk value, denoted as F.

[0023] Step 4: Obtain the path length, denoted as G, and normalize it along with the road saturation value and total risk value. Then, calculate the path priority index BF using a formula. The specific formula is as follows:

[0024] Where β1 and β2 are the set weight constants; thus, the priority index of several paths can be obtained, and the path with the largest priority index is selected as the target path; there are several inspection devices in an idle state, each inspection device corresponds to an inspection value, the inspection device with the largest inspection value is selected as the target inspection device, and the target path is assigned to the target inspection device, thereby changing the state of the target inspection device from idle state to inspection state; when the target inspection device completes the inspection task of this path, the historical inspection count of the inspection device is increased by one, and the state of the target inspection device is changed from inspection state to idle state;

[0025] Step 5: Obtain the historical number of inspections for each inspection device, and analyze the efficiency of the inspection device accordingly to obtain the inspection value of the inspection device; thus, the inspection value of each inspection device can be obtained and updated to Step 4.

[0026] Step 6: Whenever there is an idle inspection device, repeat steps 2 to 5 until all paths are inspected.

[0027] Preferably, the specific process for analyzing the saturation level of traffic flow on the route is as follows:

[0028] Choose any path and divide it into several segments. Real-time traffic flow data for each segment is acquired and compared with a set traffic flow range. When the traffic flow exceeds the maximum value within the range, the segment is designated as a highly saturated segment; when the traffic flow falls within the range, it is designated as a moderately saturated segment; and when the traffic flow is less than the minimum value, it is designated as a poorly saturated segment. The number of highly saturated, moderately saturated, and poorly saturated segments is counted and denoted as B1, B2, and B3, respectively. The traffic flows corresponding to these segments are summed to obtain high-saturation, moderate-saturation, and low-saturation traffic values, which are denoted as B4, B5, and B6, respectively.

[0029] The number of highly saturated road sections (B1), moderately saturated road sections (B2), and lowly saturated road sections (B3), as well as the high saturation flow rate (B4), moderately saturated flow rate (B5), and low saturation flow rate (B6), are normalized and their values ​​are taken. The path saturation value (Bα) is then calculated using a formula. The specific calculation formula is as follows:

[0030] α1, α2 and α3 are the set weight constants, and α1>α2>α3>1.

[0031] Preferably, the specific process for analyzing the efficiency of the inspection equipment is as follows:

[0032] Obtain the historical inspection count for each inspection device, along with the corresponding path, total risk value F, path length G, inspection start time (the moment of transition from idle to inspection state), and inspection end time (the moment of transition from inspection to idle state). Calculate the inspection duration by taking the time difference between the inspection start and end times; denoted as T, the inspection duration for each inspection is then calculated. Normalize the total risk value F, path length G, and inspection duration T, and use a formula to calculate the efficiency value for each inspection. The specific calculation formula is as follows:

[0033] Where λ1 and λ2 are the set weight constants;

[0034] Average efficiency The interval distance A is normalized and its value is taken. The value is then calculated and analyzed using a formula to obtain the inspection value Aγ of the inspection equipment. The specific calculation formula is as follows:

[0035] Where γ1 and γ2 are the set weight constants.

[0036] The beneficial effects of this invention are:

[0037] 1. By retrieving various types of defects from the central server, several inspection routes, including main roads, are automatically generated to ensure that all road defects are distributed across different routes, achieving comprehensive coverage. Traffic flow and road defects on each route are comprehensively analyzed to obtain a priority index, and the efficiency of inspection equipment is analyzed to obtain inspection values. Based on this, route inspection tasks are allocated, forming a dynamic planning of route inspection. This ensures the flexibility and adaptability of inspection activities, significantly improving the overall efficiency and safety of road inspection.

[0038] 2. By identifying road defects in the collected road images, the types and degrees of defects in the road images can be accurately extracted, providing a quantitative evaluation of defects. Furthermore, by deduplicating road images with defects and calculating the duplication index, the similarity between the target image and the comparison image can be accurately judged, avoiding duplicate detection, improving the efficiency and accuracy of defect detection, and providing intelligent and scientific decision support for road maintenance and management. Attached Figure Description

[0039] The invention will now be further described with reference to the accompanying drawings.

[0040] Figure 1 is a schematic diagram of the system module connection of the present invention;

[0041] Figure 2 is a schematic diagram of the road defect identification and deduplication process of the present invention. Detailed Implementation

[0042] 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.

[0043] Please refer to Figure 1. The present invention is a road defect inspection data analysis system based on the Internet of Things and artificial intelligence, including: a central server, an inspection planning module and a defect identification module;

[0044] The central server interacts with inspection equipment (such as inspection vehicles or drones) via a communication network. Various sensors on the inspection equipment (such as high-precision GPS and high-definition cameras) inspect along corresponding paths to collect real-time road images and their corresponding location information (location information refers to the road name and coordinates of the road image) and uploads it to the central server for storage. In addition, the central server maintains several sets of road defect types. Each set stores images of several road defects corresponding to each type, along with their location information, three key features, defect parameters, and defect severity values. The central server also communicates with the mobile devices of road maintenance personnel, sending information about each road defect, its location, and defect type to their mobile devices. When a maintenance is completed, the defect is removed from the corresponding defect type set.

[0045] The inspection planning module plans inspection paths based on the collection of various types of defects in the central server to guide inspection equipment in inspection and road image acquisition, thereby effectively improving inspection efficiency; specifically:

[0046] Step 1: Retrieve the disease images corresponding to each road disease type stored in each disease type set, along with the location information and disease severity value of each disease image. Mark the corresponding disease for each disease image at the corresponding location on the city map. Generate several paths based on the path algorithm. These paths typically include the main roads in the entire city that require road inspection. In this way, all road diseases are distributed to various paths.

[0047] Step Two: Select any route and divide it into several segments. Obtain the real-time traffic flow of each segment and compare it with a set traffic flow range. When the traffic flow exceeds the maximum value within the set range, the segment is designated as a highly saturated segment; when the traffic flow falls within the set range, it is designated as a moderately saturated segment; and when the traffic flow is less than the minimum value within the set range, it is designated as a poorly saturated segment. Count the number of highly saturated, moderately saturated, and poorly saturated segments along the route. The quantities are calculated and denoted as B1, B2, and B3 respectively. The traffic flows corresponding to highly saturated, moderately saturated, and lowly saturated road sections are summed to obtain high-saturation, moderate-saturation, and low-saturation flow values, which are denoted as B4, B5, and B6 respectively. The quantities of highly saturated road sections B1, moderately saturated road sections B2, lowly saturated road sections B3, high-saturation flow values ​​B4, moderate-saturation flow values ​​B5, and low-saturation flow values ​​B6 are normalized and their values ​​are taken. The values ​​are then used to calculate the path saturation value Bα using a formula. The specific calculation formula is as follows:

[0048] α1, α2, and α3 are set weight constants, whose values ​​are customizable and satisfy α1>α2>α3>1. As can be seen from the formula, the greater the traffic flow in each segment of the path, the greater the path saturation value, indicating that it is less suitable to carry out inspections at this time, as it will interfere with the normal operation of vehicles.

[0049] Step 3: Retrieve the number of road defects along the path, as well as the defect type and severity value for each defect; set a base value for each type of defect. Note that different types of defects have different base values: potholes have a higher base value than ruts, and ruts have a higher base value than cracks; match each road defect along the path with all the set defect types to obtain the corresponding base value; normalize the base values ​​and severity values ​​and take their numerical values; perform a weighted calculation on the numerical values ​​to obtain the defect risk value for each road defect; sum the defect risk values ​​of all road defects along the path to obtain the total risk value, denoted as F.

[0050] Step 4: Obtain the path length, denoted as G, and normalize it along with the path saturation value BQ and the total risk value F. Then, calculate the path priority index BF using a formula. The specific formula is as follows:

[0051] Where β1 and β2 are set weight constants, the magnitude of which is customized. As shown in the formula, the shorter the path length and the greater the total risk value, the greater the priority index of the path; the greater the path saturation value, the smaller the priority index of the path. Thus, several priority indices of paths can be obtained, and the path with the largest priority index is selected as the target path. Several inspection devices are set to be in an idle state, and each inspection device corresponds to an inspection value. The inspection device with the largest inspection value is selected as the target inspection device, and the target path is assigned to the target inspection device, thereby changing the status of the target inspection device from idle state to inspection state. When the target inspection device completes the inspection task of this path, the historical inspection count of the inspection device is increased by one, and the status of the target inspection device is changed from inspection state to idle state.

[0052] Step 5: Obtain the historical inspection count for each inspection device, as well as the corresponding path, total risk value F, path length G, inspection start time (the moment of transition from idle to inspection state), and inspection end time (the moment of transition from inspection state to idle state) for each inspection. Calculate the inspection duration by taking the time difference between the inspection start time and the inspection end time; denoted as T, the inspection duration for each inspection is then calculated. Normalize the total risk value F, path length G, and inspection duration T, and use a formula to calculate the efficiency value for each inspection. The specific calculation formula is as follows:

[0053] Where λ1 and λ2 are the set weight constants, the formula shows that the longer the path, the greater the risk value, and the shorter the inspection time, the higher the inspection efficiency, and thus the greater the efficiency value. The average efficiency value of the inspection equipment is calculated by averaging the efficiency values ​​of each inspection and denoted as . The interval distance between the inspection equipment in each idle state and the target path is recorded as A; the average efficiency is calculated. The interval distance A is normalized and its value is taken. The value is then calculated and analyzed using a formula to obtain the inspection value Aγ of the inspection equipment. The specific calculation formula is as follows:

[0054] Where γ1 and γ2 are the set weight constants, the inspection values ​​of each inspection device can be obtained, and they are updated to step four.

[0055] Step 6: Whenever there is an idle inspection device, repeat steps 2 to 5 until all paths are inspected;

[0056] By retrieving various types of road defects from the central server, several inspection routes, including main roads, are automatically generated to ensure that all road defects are distributed across different routes, achieving comprehensive coverage. A comprehensive analysis of traffic flow and road defects along each route is conducted to obtain a priority index, and the efficiency of the inspection equipment is analyzed to obtain inspection values. Based on this, route inspection tasks are allocated, forming a dynamic planning system for route inspection. This ensures the flexibility and adaptability of inspection activities, significantly improving the overall efficiency and safety of road inspections.

[0057] As shown in Figure 2, the road defect identification module extracts and analyzes features from road images to determine the presence of road defects and their corresponding types and severity. This allows for the identification of various road defects along different paths, and the module performs deduplication detection by comparing these defects with those stored on the server to prevent duplicate labeling. Specifically:

[0058] Road defect identification: The acquired road images are preprocessed, such as denoising, grayscale conversion, and contrast enhancement, to improve image quality. Common methods include Gaussian filtering and histogram equalization to enhance the details of the road images, facilitating the subsequent extraction and analysis of defect features. The location information of the images is parsed and associated, including road names, GPS coordinates, timestamps, etc., to ensure that each road image corresponds accurately to its corresponding location information. Geometric features (usually using edge detection, morphological operations, etc.), texture features (usually using Local Binary Pattern (LBP) or Gray-Level Co-occurrence Matrix (GLCM) algorithms), and color features (usually using color histograms or color space conversion) are extracted from the road images to obtain the three key features of the road images, which include geometric features, texture features, and color features.

[0059] Based on a labeled sample dataset, a deep learning model is trained using convolutional neural networks (CNNs) to identify different types of defects (such as cracks, potholes, ruts, and settlement). The three key features are input into the trained CNN model to output the defect type and its corresponding parameters. Different defect types correspond to different parameters; for example, cracks are represented by the number of cracks and the area of ​​each crack (the product of its length and width); potholes are represented by their diameter, depth, and edge damage; and settlement is represented by its area and depth. Each defect type is assigned a weighting coefficient, denoted as Pi, and the defect parameters are denoted as Wi. The defect severity value S is calculated by weighting the parameters according to the corresponding weighting coefficients. The specific calculation formula is as follows:

[0060] Where i = 1, 2, 3...I, I is a positive integer, I is the total number of disease parameters in a certain disease type, and i is any one of the disease parameters; thus, disease identification can be performed on each road image to filter out road images with road defects, and output the disease type and disease severity value corresponding to each road image with road defects.

[0061] Deduplication identification: Randomly select one of the road images with road defects as the target image, and use each defect image in the corresponding defect type set as the comparison image; retrieve the geometric contours of the target image and the comparison image, use the centroid calculation method to obtain the center point of the geometric contour, align the center points of the geometric contours of the target image and the comparison image, and rotate them 360° until the overlap area is the largest, then stop rotating. Calculate the intersection area and union area of ​​the overlap shape (the overlap shape with the largest overlap area), and denote them as Q1 and Q2 respectively. It should be noted that the intersection area refers to the area of ​​overlap of the geometric contours of the target image and the comparison image under the overlap shape, and the union area refers to the total area of ​​the geometric contours of the target image and the comparison image under the overlap shape.

[0062] By matching the corresponding disease parameters of the target image and the comparison image one by one, several parameter pairs can be obtained. The difference between the parameters of each parameter pair is calculated to obtain the parameter difference between the two pairs. For example, the pit diameter in the target image and the pit diameter in the comparison image are a pair of parameters, and the difference between the two is calculated to obtain the pit diameter difference. The mean of the parameter difference between the target image and the comparison image is calculated by averaging the parameter differences of each parameter pair, and it is denoted as C.

[0063] Retrieve the positional information corresponding to the target image and the comparison image, and calculate the spatial distance between them, denoted as H.

[0064] Retrieve the disease severity values ​​corresponding to the target image and the comparison image, and record them as S1 and S2 respectively;

[0065] The intersection area Q1, union area Q2, mean parameter difference C, spatial spacing H, and the severity values ​​S1 and S2 of the target image and the comparison image are normalized and their values ​​are taken. The repetition index SQ is then calculated using a formula. The specific calculation formula is as follows:

[0066] Where η1, η2, η3, and η4 are the set weight constants; it should be noted that when the intersection area between the target image and the comparison image is closer to the union area, it indicates that the geometric contours of the target image and the comparison image are closer, and the repetition index is larger; when the mean difference of parameters is smaller, it indicates that the disease parameters between the target image and the comparison image are closer, and the repetition index is larger; when the corresponding disease severity values ​​of the target image and the comparison image are closer, it indicates that the disease severity of the target image and the comparison image is closer, and the repetition index is larger; thus, the repetition index between the target image and each image in the corresponding disease type set can be obtained, and the repetition index can be compared with... The system compares and analyzes the set repetition thresholds. When the repetition index is greater than the set repetition threshold, a repetition image is counted once. When the cumulative number of repetitions is greater than or equal to one, it means that this road image has been collected in a previous inspection, and it is deleted. When the cumulative number of repetitions is zero, the road image with road defects is recorded as a defect image, and it, along with its corresponding location information, three key features, defect parameters, and defect severity value, is stored in the corresponding defect type set on the central server. Based on the dynamic update function of the inspection data, the real-time and accuracy of defect data in the central server are ensured.

[0067] By identifying road defects in the collected images, the types and severity of defects can be accurately extracted, providing a quantitative assessment of defects. Furthermore, by deduplicating road images with defects and calculating the duplication index, the similarity between the target image and the comparison image can be accurately determined, avoiding duplicate detection, improving the efficiency and accuracy of defect detection, and providing intelligent and scientific decision support for road maintenance and management.

[0068] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

Claims

1. A road disease inspection data analysis system based on Internet of Things and artificial intelligence, comprising a central server, characterized in that, Also includes: Inspection planning module and disease identification module; The inspection planning module plans inspection paths based on various types of defects to guide inspection equipment in carrying out inspections and collecting road images. The road defect identification module extracts features from road images and identifies defects to determine their presence. It then performs deduplication on road images with defects. The deduplication process involves: arbitrarily selecting one road image from those with defects as the target image; using images from the corresponding defect type set as comparison images; retrieving the geometric contours of both the target and comparison images and finding their center points; aligning the center points of the geometric contours of the target and comparison images; rotating them 360° until the overlap area is maximized; and calculating the intersection and union areas of this overlap. Finally, matching the defect parameters of the target and comparison images one by one yields several parameter pairs, and calculating the parameter difference between each pair. The mean difference between the parameters of the parameter pairs is calculated to obtain the mean parameter difference between the target image and the comparison image; the location information corresponding to the target image and the comparison image is retrieved, and the spatial distance between them is calculated accordingly; the severity values ​​of the corresponding defects in the target image and the comparison image are retrieved; the intersection area, union area, mean parameter difference, spatial distance, severity values ​​of the target image and the comparison image are analyzed to obtain the repetition index; when the repetition index is greater than the set repetition threshold, the image is counted as a repetition; when the cumulative number of repetitions is greater than or equal to one, it is deleted; when the cumulative number of repetitions is zero, the road image with defects is recorded as a defect image, and it, along with the corresponding location information, three feature elements, defect parameters, and severity values, is stored in the corresponding defect type set on the central server. 2.The IoT and artificial intelligence based road disease inspection data analysis system according to claim 1, wherein, The specific process for disease identification is as follows: The acquired road images are preprocessed to parse and associate the image location information, so that each road image is accurately matched with its corresponding location information. Geometric features, texture features and color features are extracted from the road images to obtain the three feature elements of the road images, which include geometric features, texture features and color features. Based on a labeled sample dataset, a convolutional neural network is trained to identify different types of road defects. The three key features are input into the trained convolutional neural network algorithm model to output the defect type and its corresponding defect parameters. A weighting coefficient is assigned to each defect parameter for each defect type, and the defect severity value is calculated by weighting the parameters according to the corresponding weighting coefficients. This allows for defect identification of various road images to filter out road images with defects and outputs the defect type and severity value for each road image with defects. 3.The IoT and artificial intelligence based road disease inspection data analysis system according to claim 1, wherein, The specific process of planning inspection routes based on various types of disease sets is as follows: Step 1: Retrieve the disease images of each disease type stored in each disease type set, as well as the location information and disease severity value of each disease image. Mark the disease corresponding to each disease image at the corresponding location on the city map. Generate several paths according to the path algorithm, so that all road diseases are distributed to each path. Step 2: Analyze the saturation level of traffic flow on the route to obtain the route saturation value; Step 3: Retrieve the number of road defects along the route, as well as the defect type and severity value for each defect; set a base value for each defect type; match each road defect along the route with all the set defect types to obtain the corresponding base value; normalize the base value and severity value and take their numerical values; perform weighted calculation on the numerical values ​​to obtain the defect risk value for each road defect; sum the defect risk values ​​of each road defect along the route to obtain the total risk value. Step 4: Obtain the path length, normalize it with the road saturation value and total risk value, and calculate the path priority index using a formula. This yields several path priority indices. Select the path with the highest priority index as the target path. Set up several inspection devices in an idle state, each corresponding to an inspection value. Select the inspection device with the highest inspection value as the target inspection device, and assign the target path to the target inspection device. This changes the target inspection device's state from idle to inspection state. When the target inspection equipment completes the inspection task for this path, the historical inspection count of the inspection equipment is increased by one, and the status of the target inspection equipment is changed from inspection status to idle status. Step 5: Obtain the historical number of inspections for each inspection device, and analyze the efficiency of the inspection device accordingly to obtain the inspection value of the inspection device; thus, the inspection value of each inspection device can be obtained and updated to Step 4. Step 6: Whenever there is an idle inspection device, repeat steps 2 to 5 until all paths are inspected. 4.The IoT and artificial intelligence-based road disease inspection data analysis system according to claim 3, wherein, The specific process for analyzing the saturation level of traffic flow on the route is as follows: Choose any path and divide it into several segments. Obtain the traffic flow of each segment in real time and compare it with the set traffic flow range to classify the segments corresponding to the traffic flow into highly saturated, moderately saturated, and lowly saturated segments. Count the number of highly saturated, moderately saturated, and lowly saturated segments in the path, and sum the traffic flow corresponding to each segment to obtain the high saturation flow value, moderate saturation flow value, and low saturation flow value. The number of highly saturated road sections, the number of moderately saturated road sections, the number of lowly saturated road sections, the high saturation flow value, the medium saturation flow value, and the low saturation flow value are normalized and their values ​​are taken. The values ​​are then used to calculate the path saturation value using a formula. 5.The IoT and artificial intelligence based road disease inspection data analysis system according to claim 3, wherein, The specific process for analyzing the efficiency of inspection equipment is as follows: Obtain the historical number of inspections for each inspection device, as well as the path, total risk value, path length, start time, and end time of each inspection. Calculate the inspection duration by taking the time difference between the start and end times of the inspection. The total risk value, path length, and inspection time of the path are normalized and their values ​​are taken. The values ​​are then used to calculate the efficiency value of each inspection. The average efficiency value of the inspection equipment is obtained by averaging the efficiency values ​​of each inspection. The interval distance between the inspection equipment in each idle state and the target path is obtained. The average efficiency value and the interval distance are normalized and their values ​​are taken. The values ​​are then calculated and analyzed using formulas to obtain the inspection value of the inspection equipment. 6.The IoT and artificial intelligence based road disease inspection data analysis system according to claim 1, wherein, The central server interacts with the inspection equipment via a communication network. Various sensors on the inspection equipment perform inspections along corresponding paths to collect road images and their corresponding location information in real time, and upload them to the central server for storage via the communication network. The central server also stores several sets of road damage types. Each set of road damage types stores several images of road damage corresponding to each type, as well as the location information, three key features, damage parameters, and damage severity values ​​of each image.

Images