Transportation airport pavement maintenance management system and method

Abstract

The invention relates to the technical field of image processing, in particular to a transportation airport pavement maintenance management system and method. According to the maintenance management method, firstly, whether pavement damage exists or not is judged in a standardized mode according to a pavement image obtained in the inspection process, and the accuracy and stability of pavement damage result judgment are improved; thirdly, carrying out image semantic segmentation processing on the pavement image containing the pavement damage to obtain a pavement damage semantic feature image which is convenient for expressing pavement damage features and is used for improving the accuracy of pavement damage type identification; then, calculating the influence degree of each pavement damage type, and selecting the pavement damage of the area corresponding to the pavement damage type of which the influence degree is greater than an influence threshold value as the pavement damage to be overhauled; then, the to-be-overhauled pavement damage is overhauled according to the sequence of the area weights from large to small, and the real-time performance, efficiency and quality of pavement damage overhauling are improved under the condition that the labor cost is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a transport airport pavement maintenance management system and method. Background Art

[0002] Airport pavement maintenance is a crucial component of airport pavement maintenance and management. Pavement damage refers to the presence of tears, cracks, potholes, and wrinkles on runways, taxiways, and aprons. Pavement damage can affect aircraft takeoff and landing; it can also subject aircraft to additional shock and vibration during takeoff and landing, increasing maintenance costs; it can also accelerate aircraft wear and tear, shortening its service life. Severe pavement damage can even cause problems during ground operations, leading to accidents.

[0003] To mitigate these hazards, airport managers typically conduct regular inspections and repairs of pavement damage. Because manual judgment is influenced by individual subjective opinions and experience, different personnel may have different assessment criteria for the severity of pavement damage, leading to inconsistent judgment results. Furthermore, manual judgment may struggle to accurately detect minor pavement damage or fine cracks, which may be early-stage issues but have significant long-term impacts on the pavement structure. Furthermore, airport pavements typically cover large areas, and manual inspections require significant human resources and time, making frequent inspections impractical for large airports. Furthermore, when a large number of pavement damage inspections need to be handled, human resources and time constraints can affect the real-time, efficiency, and quality of pavement damage repairs. Summary of the Invention

[0004] The purpose of the present invention is to provide a transport airport pavement maintenance management system and method that can save labor costs and improve maintenance quality and efficiency.

[0005] The technical solutions of the present invention are as follows: A method for maintaining and managing a transport airport pavement includes the following operations: S1. Inspect the pavement of the transport airport according to a preset route; based on the obtained pavement map of the current position, determine whether the pavement at the current position is damaged; if so, execute S2; if not, inspect the pavement at the next position according to the preset route; S2. Processing the current position road surface image through image semantic segmentation to obtain a current road surface damage semantic feature map; obtaining similarity between the current road surface damage semantic feature map and each standard road surface damage map in a standard road surface damage database, and determining the road surface damage type of the standard road surface damage map corresponding to the maximum similarity as the current road surface damage type; S3. After completing one inspection along the preset route, all pavement damage types are obtained to form a pavement damage type set; based on the pavement damage area, the influence of each pavement damage type in the pavement damage type set is obtained; and the pavement damage corresponding to the pavement damage type with an influence greater than the influence threshold is repaired in order of regional weight from large to small.

[0006] The operation of determining whether there is pavement damage on the pavement at the current position based on the acquired pavement map at the current position in S1 specifically comprises: determining whether there is a crack on the pavement at the current position based on the acquired pavement map at the current position; If it exists, obtain the crack width in the road surface map at the current position; if the crack width is greater than the width threshold, there is road surface damage on the road surface at the current position; if the crack width is not greater than the width threshold, obtain the crack ratio of the road surface at the current position; if the crack ratio exceeds the proportion threshold, there is road surface damage on the road surface at the current position.

[0007] The operation of determining whether there is a crack on the road surface at the current position based on the acquired road surface map at the current position is specifically as follows: grayscale processing is performed on the road surface map at the current position to obtain a current road surface grayscale map; grayscale values of all position points in the current road surface grayscale map are obtained to obtain a current road surface grayscale distribution map; and determining whether there is an area in the current road surface grayscale distribution map whose average grayscale value is greater than a grayscale threshold and whose area is greater than an area threshold; if so, cracks exist on the road surface at the current position; if not, cracks do not exist on the road surface at the current position.

[0008] The operation of obtaining the crack ratio is specifically as follows: obtaining the pixel value of each position point in the current position road surface map to obtain the current position road surface pixel value distribution map; marking the points whose pixel values exceed the pixel threshold in the current position road surface pixel value distribution map as black, and marking the points whose pixel values do not exceed the pixel threshold as white, to obtain the current road surface marking map; after morphological processing of the current road surface marking map, calculating the ratio of the number of black pixels to the number of white pixels to obtain the crack ratio.

[0009] Before the operation of obtaining the pixel value distribution map of the road surface at the current position, the process also includes performing color correction processing on the road surface map at the current position; the color correction processing operation is specifically as follows: obtaining the color temperature difference between the color temperature of the road surface map at the current position and the standard color temperature, and based on the color temperature difference, obtaining the channel gain of different color channels of each position point in the road surface map at the current position to obtain the red gain, green gain and blue gain of each position point; obtaining the white balance coefficient of each position point based on the red gain, green gain and blue gain of each position point; dividing the pixel value of each position point in the road surface map at the current position by the corresponding white balance coefficient to obtain an updated pixel value of each position point, and obtaining the current color correction map; the current color correction map is used to perform the operation of obtaining the pixel value distribution map of the road surface at the current position.

[0010] The operations of the image semantic segmentation processing in S2 are specifically as follows: the road surface map at the current position is convolved and down-sampled to obtain a first down-sampled feature image; the first down-sampled feature image is convolved and down-sampled to obtain a second down-sampled feature image; the second down-sampled feature image is convolved and down-sampled to obtain a third down-sampled feature image; the third down-sampled feature image is convolved and up-sampled, and then fused with the third down-sampled feature image to obtain a third fused feature image; the third fused feature image is convolved and up-sampled, and then fused with the second down-sampled feature image to obtain a second fused feature image; the second fused feature image is convolved and up-sampled, and then fused with the first down-sampled feature image to obtain the current road surface damage semantic feature map.

[0011] The third down-sampled feature image, and / or the third fused feature image, and / or the second fused feature image, after convolution and upsampling, are processed by the attention mechanism, and the obtained third attention feature map, and / or the second attention feature map, and / or the first attention feature map are respectively spliced with the third down-sampled feature image, and / or the second down-sampled feature image, and / or the first down-sampled feature image.

[0012] In S3, the influence of each road surface damage type is obtained by the total road surface damage area, the total number of road surface damage sub-areas and the area weight of each corresponding region.

[0013] A transport airport pavement maintenance management system, including a web terminal and an APP terminal: The APP is used to inspect the pavement of the transport airport according to a preset route and transmit the obtained pavement map of the current location to the web terminal; it is used to obtain the impact of each pavement damage type on the web terminal, and repair the pavement damage corresponding to the pavement damage type with an impact greater than the impact threshold in order of regional weight from large to small; The web end is used to determine whether there is pavement damage on the road at the current position based on the current position pavement map obtained by the APP end; if not, prompt the APP end to inspect the next position road according to the preset route; if so, perform image semantic segmentation processing on the current position road map to obtain a current pavement damage semantic feature map; obtain the similarity between the current pavement damage semantic feature map and each standard pavement damage map in the standard pavement damage database, and use the pavement damage type of the standard pavement damage map corresponding to the maximum similarity as the current pavement damage type; obtain all pavement damage types obtained after the APP end completes an inspection along the preset route once, forming a pavement damage type set; based on the pavement damage area, obtain the influence of each pavement damage type in the pavement damage type set.

[0014] The APP terminal inspects the airport pavement according to the preset route and realizes positioning by connecting to the real-time motion positioning system connected to the positioning base station through Bluetooth.

[0015] The beneficial effects of the present invention are: The present invention provides a transport airport pavement maintenance and management method. First, based on the pavement images obtained during the inspection process, a standardized judgment is made on whether pavement damage exists, thereby improving the accuracy and stability of the pavement damage result judgment; then, the pavement image containing the pavement damage is subjected to image semantic segmentation processing to obtain a pavement damage semantic feature map that is convenient for expressing the pavement damage characteristics, which is used to improve the accuracy of pavement damage type identification; then, the influence of each pavement damage type is calculated, and the pavement damage in the area corresponding to the pavement damage type with an influence greater than an influence threshold is selected as the pavement damage to be repaired; then, the pavement damage to be repaired is repaired in descending order of regional weights. This method improves the real-time performance, efficiency and quality of pavement damage repair while saving labor costs. DETAILED DESCRIPTION

[0016] This embodiment provides a method for maintaining and managing a transport airport pavement, including the following operations: S1. Inspect the pavement of the transport airport according to a preset route; based on the obtained pavement map of the current position, determine whether the pavement at the current position is damaged; if so, execute S2; if not, inspect the pavement at the next position according to the preset route; S2. Processing the current position road surface image through image semantic segmentation to obtain a current road surface damage semantic feature map; obtaining similarity between the current road surface damage semantic feature map and each standard road surface damage map in a standard road surface damage database, and determining the road surface damage type of the standard road surface damage map corresponding to the maximum similarity as the current road surface damage type; S3. After completing one inspection along the preset route, all pavement damage types are obtained to form a pavement damage type set; based on the pavement damage area, the influence of each pavement damage type in the pavement damage type set is obtained; and the pavement damage corresponding to the pavement damage type with an influence greater than the influence threshold is repaired in order of regional weight from large to small.

[0017] S1. Inspect the transport airport pavement according to the preset route; based on the obtained current position pavement map, determine whether there is pavement damage on the current position pavement; if so, execute S2; if not, inspect the next position pavement according to the preset route.

[0018] Based on the layout of the transport airport pavement, an inspection route is independently designed and the transport airport pavement is inspected according to the preset route. During the inspection process, based on the obtained pavement map of the current location, it is first determined whether the pavement at the current location is damaged. If the pavement at the current location is damaged, S2 is executed to determine the type of pavement damage. If the pavement at the current location is not damaged, the next location is inspected according to the preset route to determine the pavement damage.

[0019] Based on the acquired road surface map at the current position, the operation for determining whether there is road surface damage on the road surface at the current position is as follows: based on the acquired road surface map at the current position, determine whether there is a crack on the road surface at the current position; if there is, obtain the crack width in the road surface map at the current position; if the crack width is greater than the width threshold, there is road surface damage on the road surface at the current position; if the crack width is not greater than the width threshold, obtain the crack ratio of the road surface at the current position; if the crack ratio exceeds the ratio threshold, there is road surface damage on the road surface at the current position.

[0020] Based on the acquired road surface map at the current position, the operation for determining whether there are cracks on the road surface at the current position is as follows: the road surface map at the current position is gray-scaled to obtain a current road surface grayscale map; the grayscale values of all position points in the current road surface grayscale map are obtained to obtain a current road surface grayscale distribution map; and it is determined whether there is an area in the current road surface grayscale distribution map with an average grayscale value greater than a grayscale threshold and an area greater than an area threshold; if so, cracks exist on the road surface at the current position; if not, cracks do not exist on the road surface at the current position.

[0021] If there are cracks on the road surface at the current location and the crack width is too large, it means that there are some holes on the road surface at the current location that are likely to cause traffic hazards and destroy the stability of the road surface. At this time, there is road surface damage on the road surface at the current location; if the crack width is small, it is impossible to get an accurate judgment result based solely on the length of the crack. Therefore, it is necessary to use the crack ratio that can reflect the current distribution of road surface cracks to judge the road surface damage.

[0022] The operation of obtaining the crack ratio is as follows: obtain the pixel value of each position point in the current position road surface map to obtain the current position road surface pixel value distribution map; mark the points with pixel values exceeding the pixel threshold as black and the points with pixel values not exceeding the pixel threshold as white in the current position road surface pixel value distribution map to obtain the current road surface marking map; after the current road surface marking map is morphologically processed, calculate the ratio of the number of black pixels to the number of white pixels to obtain the crack ratio.

[0023] Specifically, the pixel points at each position on the current position road surface map are extracted to obtain the current position road surface pixel value distribution map; then, in the current position road surface pixel value distribution map, the points whose pixel values exceed the pixel threshold are marked as black, that is, the positions corresponding to the cracks are marked as black, and the points whose pixel values do not exceed the pixel threshold are marked as white, that is, the positions corresponding to the non-cracks are marked as white, to obtain the current road surface marking map; subsequently, the current road surface marking map is subjected to image opening operation (first corrosion and then expansion) to realize morphological processing, eliminate some error points in the current road surface marking map that are far away from the crack area, and remove image noise; finally, the ratio of the number of black pixels to the number of white pixels in the current road surface marking map after morphological processing is used as the crack ratio that can reflect the density of crack distribution.

[0024] When obtaining the road surface map at the current location, due to the influence of weather, the color of the road surface map at the current location may be generally dark, which is not conducive to the expression of crack characteristics and may reduce the accuracy of the crack ratio, thereby affecting the judgment of road surface damage.

[0025] To solve this technical problem, before obtaining the pixel value distribution map of the road surface at the current position, the road surface map at the current position is also color corrected. Color correction can eliminate the hue deviation in the road surface map at the current position, making the road surface map at the current position more consistent with the color under real lighting conditions.

[0026] The specific operations of the color correction processing are as follows: obtaining the color temperature difference between the color temperature of the road surface image at the current position and the standard color temperature, and based on the color temperature difference, obtaining the channel gain of different color channels of each position point in the road surface image at the current position to obtain the red gain, green gain and blue gain of each position point; based on the red gain, green gain and blue gain of each position point, obtaining the white balance coefficient of each position point; dividing the pixel value of each position point in the current road surface crack image by the corresponding white balance coefficient to obtain the updated pixel value of each position point, implement color correction processing, and obtain the current color correction image; the current color correction image is used to execute the operation of obtaining the pixel value distribution map of the road surface at the current position.

[0027] S2. The current position road surface image is processed by image semantic segmentation to obtain the current road surface damage semantic feature map; the similarity between the current road surface damage semantic feature map and each standard road surface damage map in the standard road surface damage database is obtained, and the road surface damage type of the standard road surface damage map corresponding to the maximum similarity is used as the current road surface damage type.

[0028] By performing image semantic segmentation on the current position road surface image where road surface damage exists, a current road surface damage semantic feature map that is convenient for expressing the characteristics of road surface damage can be obtained, thereby improving the accuracy of subsequent comparative recognition and classification.

[0029] The operation of image semantic segmentation processing is as follows: the current position road surface map is convolved and down-sampled to obtain a first down-sampled feature image; the first down-sampled feature image is convolved and down-sampled to obtain a second down-sampled feature image; the second down-sampled feature image is convolved and down-sampled to obtain a third down-sampled feature image; the third down-sampled feature image is convolved and up-sampled, and then fused with the third down-sampled feature image to obtain a third fused feature image; the third fused feature image is convolved and up-sampled, and then fused with the second down-sampled feature image to obtain a second fused feature image; the second fused feature image is convolved and up-sampled, and then fused with the first down-sampled feature image to obtain a current road surface damage semantic feature map.

[0030] The number of convolution, downsampling, and upsampling processing times can be adjusted according to actual needs, including but not limited to the above-mentioned combination of three convolutions and downsampling, and the combination of three convolutions and upsampling.

[0031] To better understand and combine the high-level and low-level semantic features of the road surface image at the current location, the algorithm focuses on crack regions and reduces useless background feature information, thereby improving the accuracy of crack feature extraction. After convolution and upsampling, the third downsampled feature image, and / or the third fused feature image, and / or the second fused feature image are processed using an attention mechanism. The resulting third attention feature map, and / or the second attention feature map, and / or the first attention feature map are then concatenated with the third downsampled feature image, and / or the second downsampled feature image, and / or the first downsampled feature image.

[0032] The operation of the attention mechanism is as follows: the feature input map processed by convolution and upsampling is processed by channel attention and spatial attention in turn, and the obtained attention feature map is used to perform splicing processing.

[0033] The operation of channel attention processing is as follows: the feature input map is processed by global maximum pooling and global average pooling respectively to obtain the first maximum pooling map and the first average pooling map; the first maximum pooling map and the first average pooling map are processed by the multi-layer perceptron to obtain the first channel weight and the second channel weight; based on the first channel weight and the second channel weight, the first maximum pooling map and the first average pooling map are weighted, and the obtained pooling fusion map is multiplied element-by-element with the feature input map to obtain the channel attention map.

[0034] The operation of spatial attention processing is as follows: the channel attention map is processed by global maximum pooling and global average pooling respectively, and the first maximum pooling channel map and the first average pooling channel map are spliced and then convolved to obtain a spatial convolution map; the spatial convolution map and the channel attention map are element-wise multiplied to obtain an attention feature map.

[0035] Before the operation of image semantic segmentation processing, detail enhancement processing is also included on the current position road surface map to obtain the current road surface detail enhancement damage map; the current road surface detail enhancement damage map is used to perform the image semantic segmentation processing operation.

[0036] The detail enhancement operation is as follows: the current position road surface image is Gaussian blurred to obtain a smooth image; the current position road surface image and the smooth image are subtracted to obtain a detail image; the detail image and the current position road surface image are superimposed to obtain the current road surface detail enhanced damage image.

[0037] Finally, the semantic feature map of the current pavement damage is obtained and compared with each standard pavement damage map in the standard pavement damage database to obtain the similarity (this can be achieved through existing neural networks, but to save space, it is not described in detail here). The pavement damage type of the standard pavement damage map corresponding to the maximum similarity value is used as the current pavement damage type, which is used to sort the degree of danger (impact) when it is to be repaired in the subsequent S3.

[0038] Taking cement concrete pavements as an example, pavement damage types are mainly divided into structural damage and non-structural damage. Structural damage includes: longitudinal, transverse, and diagonal cracks; corner fractures; broken slabs or cross cracks; subsidence or misalignment; and expansion cracks. Non-structural damage includes: caulking material damage, joint shattering, durability cracks, shrinkage cracks, peeling, crazing, fine cracks, and flaking of slab corners.

[0039] S3. After completing an inspection along the preset route, all pavement damage types are obtained to form a pavement damage type set; based on the pavement damage area, the impact of each pavement damage type in the pavement damage type set is obtained; the pavement damage corresponding to the pavement damage type with an impact greater than the impact threshold is repaired in order of regional weight from large to small.

[0040] After inspecting the transport airport pavement along the preset route, all pavement damage types are obtained, forming a pavement damage type set. The impact of each pavement damage type in the pavement damage type set is calculated, and the pavement damage in the area corresponding to the pavement damage type with an impact greater than the impact threshold is selected as the pavement damage to be repaired. Next, the pavement damage to be repaired is repaired in descending order of regional weight. If multiple pavement damage types with an impact greater than the impact threshold occur in the same repair area, the pavement damage corresponding to each type is repaired in descending order of impact.

[0041] The impact of each pavement damage type is obtained by the total pavement damage area, the total number of pavement damage sub-areas and the regional weight of each corresponding region.

[0042] The influence can be obtained by the following formula: , I k For the k The impact of various road surface damage types; W i For transport airport pavement i Regional weight of each region; S ik For transport airport pavement i In the region k The total pavement damage area of the first pavement damage type, i.e. i In the region k The sum of the pavement damage areas of all sub-regions for each pavement damage type, i The value of is [1, M ], M is the total number of areas divided into which the transport airport pavement is located; the pavement damage area of the current sub-area, that is, the pavement damage area at the current location, is the area of the circumscribed circle or rectangle of the crack in the pavement map at the current location; S is the total pavement damage area, which is the sum of the areas of all pavement damages; C 1 It is the area influence parameter, and its value is greater than 1; N ik For transport airport pavement i In the region k The total number of pavement damage sub-areas of the pavement damage type, i.e. i There is a k The total number of sub-areas for each pavement damage type, k The value of is [1, K ],K is the total number of pavement damage types; N is the total number of pavement damages, which is the sum of the number of all pavement damages; C 2 It is a quantity influence parameter, and its value is greater than 1.

[0043] This embodiment also provides a transport airport pavement maintenance management system, including a web side and an APP side.

[0044] The APP end is used to inspect the pavement of the transport airport according to the preset route and transmit the obtained pavement map of the current position to the web end; it is used to obtain the impact of each pavement damage type in the web end, and the pavement damage corresponding to the pavement damage type with an impact greater than the impact threshold will be inspected in order from large to small according to the regional weight.

[0045] The app is designed for inspectors and maintenance personnel, primarily responsible for pavement inspection and maintenance. Its main functions include base station setup, inspection, maintenance, data query, and other account operations. Inspection functions include recording images of pavement damage; maintenance functions include querying repair addresses and entering repair information. Once a repair is completed, it is submitted to the web for review. If approved, the repair is complete; if not, it is returned for further repair. Data query allows viewing the distribution of pavement damage.

[0046] The app is used to inspect the pavement of transport airports according to preset routes. It achieves communication positioning through Bluetooth connection with a real-time kinematic system (RTK) connected to a positioning base station.

[0047] The web side is used to determine whether there is pavement damage on the current position road surface based on the current position road surface map obtained by the APP side; if not, the APP side is prompted to inspect the next position road surface according to the preset route; if so, the current position road surface map is subjected to image semantic segmentation processing to obtain the current road surface damage semantic feature map; the current road surface damage semantic feature map is obtained, and the similarity with each standard road surface damage map in the standard road surface damage database is obtained, and the pavement damage type of the standard road surface damage map corresponding to the maximum similarity is used as the current road surface damage type; it is used to obtain all pavement damage types obtained after the APP side completes an inspection according to the preset route, forming a pavement damage type set; based on the pavement damage area, the influence of each pavement damage type in the pavement damage type set is obtained.

[0048] The web terminal includes a web client and a web management terminal. The web client is intended for inspectors, maintenance personnel, and management personnel, and is primarily responsible for obtaining and counting pavement damage types, damaged areas, and quantities, as well as impacts, and for querying historical maintenance records. The web management terminal is intended for system administrators, and is primarily responsible for managing pavement project startups and shutdowns, account and permission configuration, process configuration, and data management. Data management includes importing coordinate and image data and exporting project data.

[0049] The system usage process includes app-based inspection and maintenance, and web-based data storage and data query and retrieval. Hardware support includes mobile phones, RTK, and positioning base stations.

[0050] First, connect RTK to the positioning base station, use the mobile phone APP to connect to RTK to obtain centimeter-level high-precision positioning information, and provide it to the system for use.

[0051] App-based inspection. Inspectors use the app to conduct road surface inspections, capture road surface images, enter inspection information, and obtain road surface damage results. If damage is detected at the current location, a repair request is sent to the web client. If the web client agrees to the repair, the repair process begins. If not, the next road surface location is inspected along the pre-set route.

[0052] The maintenance process includes both pending and rejected repairs: maintenance personnel accept a maintenance task, select the task, and then enter the navigation system. They are guided to the damaged pavement to be repaired and perform the pending inspection. When the pending inspection is complete, they fill in the corresponding information (including uploading the post-inspection image on the web client), completing the pending inspection task. The inspection results are reviewed by the web client to determine if they are qualified. If qualified, a maintenance record is generated and archived. If unqualified, the maintenance personnel continue the inspection.

[0053] Web-side data storage. Road surface data is entered into the app and stored in the web client's server database. The server database is controlled to implement data storage, indexing, querying, access control, data backup and recovery, and other data control functions.

[0054] Data control includes: table structure management, which is responsible for managing and maintaining the table structure of data, including the selection and creation, modification and deletion of fields; index management, which is responsible for managing and maintaining the index structure of data, including the creation, modification and deletion of indexes; query processing, which is responsible for managing and maintaining query operations in the table, including selection, calculation, sorting and filtering; data integrity management, which is responsible for managing and maintaining data integrity, including access control, verification and validation, to ensure the integrity of data during transmission, storage and processing; data backup and recovery, which is responsible for managing and maintaining data backup and recovery, to ensure the security and reliability of data during transmission, storage and processing.

[0055] The system's data includes pavement damage type data, pavement damage area data, pavement damage quantity data, impact data, pavement SCI data, and pavement PCI data. The data in the system is visual.

[0056] The pavement damage type data is included in the design drawings. The design drawings and map data are fused to generate a map, which is imported into the system to provide a base map for data display, with multiple layers (regions, units, plates) displayed and containing coordinate information.

[0057] The system displays the pavement damage distribution, pavement damage type distribution, impact distribution, pavement SCI and pavement PCI distribution, and pavement structure in the form of a map; and displays the pavement damage area statistics, pavement damage quantity statistics, impact statistics, pavement SCI and pavement PCI statistics, maintenance number statistics, and material consumption statistics (repair materials) in the form of a chart.

[0058] Pavement damage distribution: Displays the pavement damage distribution in the form of maps and lists.

[0059] Pavement damage information is divided into pavement damage awaiting review, pavement damage awaiting repair, and pavement damage that has been repaired. Detailed information including pictures and videos showing the corresponding pavement damage is provided, and the severity of the pavement damage is indicated.

[0060] Influence Distribution: Displays the influence distribution in the form of maps and lists. Different levels of influence are marked with colors, and status charts (histograms, etc.) can be automatically generated.

[0061] Pavement Structure: Based on the query results, the airport pavement area and zoning structure are displayed. Pavement structure information includes area, construction time, zoning structure, and area weight value.

[0062] Report generation: Generate a pavement condition report based on system data AI, including the above-mentioned pavement condition data statistics (damage area statistics, pavement damage quantity statistics, impact statistics, maintenance number statistics, material consumption statistics, etc.) and various indicators (impact, pavement SCI and pavement PCI) distribution maps, and various indicator graphic reports.

[0063] This embodiment provides a transport airport pavement maintenance and management method. First, based on the pavement images obtained during the inspection process, a standardized judgment is made on whether pavement damage exists, thereby improving the accuracy and stability of the pavement damage result judgment; then, the pavement image containing the pavement damage is subjected to image semantic segmentation processing to obtain a pavement damage semantic feature map that is convenient for expressing the pavement damage characteristics, which is used to improve the accuracy of pavement damage type identification; then, the influence of each pavement damage type is calculated, and the pavement damage in the area corresponding to the pavement damage type with a larger influence is selected as the pavement damage to be repaired; finally, the pavement damage to be repaired in the corresponding area with a larger area weight is given priority repair. This method saves labor costs while improving the real-time, efficiency and quality of pavement damage repair.

Claims

1. A method for maintaining and managing a transport airport pavement, characterized in that: The following operations are included: S1. Inspect the pavement of the transport airport according to a preset route; based on the obtained pavement map of the current position, determine whether the pavement at the current position is damaged; if so, execute S2; if not, inspect the pavement at the next position according to the preset route; S2. Processing the current position road surface image through image semantic segmentation to obtain a current road surface damage semantic feature map; obtaining similarity between the current road surface damage semantic feature map and each standard road surface damage map in a standard road surface damage database, and determining the road surface damage type of the standard road surface damage map corresponding to the maximum similarity as the current road surface damage type; S3. After completing one inspection along the preset route, all pavement damage types obtained form a pavement damage type set; Based on the pavement damage area, obtaining the influence of each pavement damage type in the pavement damage type set; The pavement damage corresponding to the pavement damage type with an impact greater than the impact threshold will be inspected and repaired in order of regional weight from large to small.

2. The method for maintaining and managing a transport airport pavement according to claim 1, characterized in that: The operation of determining whether there is road surface damage on the road surface at the current position based on the acquired road surface map at the current position in S1 is specifically as follows: Based on the acquired road surface map at the current position, determine whether there is a crack on the road surface at the current position; If it exists, obtain the crack width in the road surface map at the current position; if the crack width is greater than the width threshold, there is road surface damage on the road surface at the current position; if the crack width is not greater than the width threshold, obtain the crack ratio of the road surface at the current position; if the crack ratio exceeds the proportion threshold, there is road surface damage on the road surface at the current position.

3. The method for maintaining and managing the transport airport pavement according to claim 2, characterized in that: The operation of determining whether there is a crack on the road surface at the current position based on the acquired road surface map at the current position is specifically as follows: The current position road surface map is grayscale processed to obtain the current road surface grayscale map; the grayscale values of all position points in the current road surface grayscale map are obtained to obtain the current road surface grayscale distribution map; it is determined whether there is an area in the current road surface grayscale distribution map with an average grayscale value greater than the grayscale threshold and an area greater than the area threshold; if so, there is a crack on the road surface at the current position; if not, there is no crack on the road surface at the current position.

4. The method for maintaining and managing the transport airport pavement according to claim 2, characterized in that: The specific operation of obtaining the crack ratio is: Obtaining the pixel value of each position point in the current position road surface map to obtain a current position road surface pixel value distribution map; In the pixel value distribution map of the road surface at the current position, points whose pixel values exceed the pixel threshold are marked black, and points whose pixel values do not exceed the pixel threshold are marked white, to obtain a current road surface marking map; After the current road surface marking image is morphologically processed, the ratio of the number of black pixels to the number of white pixels is calculated to obtain the crack ratio.

5. The method for maintaining and managing the transport airport pavement according to claim 4, characterized in that: Before obtaining the pixel value distribution map of the road surface at the current position, the process also includes performing color correction processing on the road surface map at the current position; The color correction processing specifically includes: obtaining a color temperature difference between a color temperature of a current position road surface image and a standard color temperature; obtaining, based on the color temperature difference, channel gains of different color channels at each position point in the current position road surface image to obtain red gain, green gain, and blue gain at each position point; obtaining a white balance coefficient at each position point based on the red gain, green gain, and blue gain at each position point; dividing a pixel value of each position point in the current position road surface image by the corresponding white balance coefficient to obtain an updated pixel value at each position point, thereby obtaining a current color correction image; The current color correction map is used to perform the operation of obtaining the road surface pixel value distribution map at the current position.

6. The method for maintaining and managing a transport airport pavement according to claim 1, characterized in that: The operations of the image semantic segmentation processing in S2 are specifically as follows: The road surface map at the current position is convolved and down-sampled to obtain a first down-sampled feature image; the first down-sampled feature image is convolved and down-sampled to obtain a second down-sampled feature image; the second down-sampled feature image is convolved and down-sampled to obtain a third down-sampled feature image; the third down-sampled feature image is convolved and up-sampled, and then fused with the third down-sampled feature image to obtain a third fused feature image; the third fused feature image is convolved and up-sampled, and then fused with the second down-sampled feature image to obtain a second fused feature image; the second fused feature image is convolved and up-sampled, and then fused with the first down-sampled feature image to obtain the current road surface damage semantic feature map.

7. The method for maintaining and managing a transport airport pavement according to claim 6, characterized in that: The third down-sampled feature image, and / or the third fused feature image, and / or the second fused feature image, after convolution and upsampling, are processed by the attention mechanism, and the obtained third attention feature map, and / or the second attention feature map, and / or the first attention feature map are respectively spliced with the third down-sampled feature image, and / or the second down-sampled feature image, and / or the first down-sampled feature image.

8. The method for maintaining and managing a transport airport pavement according to claim 1, characterized in that: In S3, the influence of each road surface damage type is obtained by the total road surface damage area, the total number of road surface damage sub-areas and the area weight of each corresponding region.

9. A transport airport pavement maintenance and management system, characterized in that: Including web and APP: The APP is used to inspect the pavement of the transport airport according to a preset route and transmit the obtained pavement map of the current location to the web terminal; it is used to obtain the impact of each pavement damage type on the web terminal, and repair the pavement damage corresponding to the pavement damage type with an impact greater than the impact threshold in order of regional weight from large to small; The web end is used to determine whether there is pavement damage on the road at the current position based on the current position pavement map obtained by the APP end; if not, prompt the APP end to inspect the next position road according to the preset route; if so, perform image semantic segmentation processing on the current position road map to obtain a current pavement damage semantic feature map; obtain the similarity between the current pavement damage semantic feature map and each standard pavement damage map in the standard pavement damage database, and use the pavement damage type of the standard pavement damage map corresponding to the maximum similarity as the current pavement damage type; obtain all pavement damage types obtained after the APP end completes an inspection along the preset route once, forming a pavement damage type set; based on the pavement damage area, obtain the influence of each pavement damage type in the pavement damage type set.

10. The transport airport pavement maintenance and management system according to claim 9, characterized in that: The APP terminal inspects the airport pavement according to the preset route and realizes positioning by connecting to the real-time motion positioning system connected to the positioning base station through Bluetooth.