Method and system for judging abnormality of elevated facility based on environmental feature fusion

Abstract

This invention discloses a method and system for anomaly detection of elevated facilities based on environmental feature fusion. The method includes: acquiring video streams, GPS location data, and timestamp information to obtain inspection data; identifying and determining specific elevated facilities and their components, and defining the target inspection area; performing dual-channel detection on specific elevated facilities based on a dynamically learned individual appearance benchmark model and an individual spatial relationship model to identify potential problems; comparing the material types and standard appearance characteristics of specific elevated facilities in each road segment with the dynamically learned individual appearance benchmark model to obtain appearance inspection results; tracking and recording the continuous spatial distribution patterns of specific elevated facilities in each road segment based on the dynamically learned individual spatial relationship model; integrating the continuous spatial distribution patterns of specific elevated facilities in each road segment with the appearance inspection results to identify potential problems; and generating an anomaly report. This invention improves the accuracy and adaptability of elevated facility detection.

Description

Technical Field

[0001] This invention relates to computer vision, and more specifically to a method and system for judging anomalies in elevated facilities based on environmental feature fusion. Background Technology

[0002] With the acceleration of urbanization, elevated infrastructure, as a key component of the modern transportation system, is becoming increasingly important. These facilities not only support the daily operation of cities but also directly relate to public safety. However, due to differences in construction period, road section, and materials, each elevated infrastructure possesses unique individual characteristics and a complex spatial distribution. Traditional detection technologies often rely on general standard libraries for comparative analysis, which proves inadequate when dealing with diverse and dynamically changing elevated infrastructure, resulting in a high false alarm rate and low adaptability.

[0003] Furthermore, elevated facilities undergo multiple renovations and upgrades throughout their lifecycle, resulting in changes to their spatial structure and appearance. Traditional static models struggle to capture these dynamic changes, impacting the accuracy and reliability of inspection results. More importantly, existing automated inspection technologies primarily focus on identifying surface damage, failing to effectively assess the overall structural order of the facility and limiting their comprehensive diagnostic capabilities.

[0004] Therefore, it is necessary to design a new method to improve the accuracy and adaptability of elevated facility inspection and provide reliable support for the digital management of elevated facilities. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and system for judging anomalies in elevated facilities based on environmental feature fusion.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for judging anomalies in elevated facilities based on environmental feature fusion, comprising:

[0007] The inspection data is obtained by acquiring video streams, GPS location data, and timestamp information collected by equipment on the dynamic inspection vehicle.

[0008] The inspection data is compared with a pre-set geographic information database of elevated facilities to identify and determine specific elevated facilities and components, and to determine the target range for inspection.

[0009] Based on a dynamic learning-based individual appearance benchmark model and an individual spatial relationship model, dual-channel detection is performed on the specific elevated facilities to identify potential problems. This includes: comparing the material types and standard appearance characteristics of the specific elevated facilities in each road segment using the dynamic learning-based individual appearance benchmark model to obtain appearance detection results; tracking and recording the continuous spatial distribution patterns of the specific elevated facilities in each road segment using the dynamic learning-based individual spatial relationship model; and integrating the continuous spatial distribution patterns of the specific elevated facilities in each road segment with the corresponding appearance detection results to identify potential problems.

[0010] An anomaly report is generated based on the potential issues.

[0011] The further technical solution is as follows: the process of establishing the individual appearance benchmark model includes:

[0012] Acquire multiple frames of images;

[0013] Analyze multiple consecutive frames of images to identify and record the specific material types and standard appearance characteristics of various facilities in different road sections in order to obtain feature data;

[0014] Based on the aforementioned feature data, an individualized appearance benchmark library is constructed for each road segment and its internal different facility types. When there are changes in the appearance of facilities or updates to materials, the individualized appearance benchmark library is updated.

[0015] The further technical solution is as follows: the process of establishing the individual spatial relationship model includes:

[0016] List all facilities with spatial continuity within the current road segment, and perform time-series tracking and trajectory fitting for each type of continuous facility to determine the standard spacing, relative position, and linear arrangement pattern of the facilities within their distribution road segment, thus generating analysis results;

[0017] Based on the analysis results, a spatial relationship benchmark database is established for different facility types and road segments, forming individual spatial relationship models. When a change in spatial relationship is detected, the corresponding spatial relationship benchmark database is updated.

[0018] The further technical solution is as follows: the individual appearance benchmark model based on dynamic learning compares the material types and standard appearance characteristics of specific elevated facilities in each road segment to obtain appearance inspection results; the individual spatial relationship model based on dynamic learning tracks and records the continuous spatial distribution pattern of specific elevated facilities in each road segment; and integrates the continuous spatial distribution pattern of the specific elevated facilities in each road segment with the corresponding appearance inspection results to identify potential problems, including:

[0019] Using the different road section facility material types and standard appearance characteristics recorded in the individual appearance benchmark model, the appearance of the specific elevated facility is analyzed to obtain the appearance inspection results;

[0020] The spatial distribution pattern and structural characteristics of the specific elevated facilities are monitored in real time and compared with the data in the individual spatial relationship model to identify abnormal changes or deviations that affect the structural integrity of the facilities, so as to obtain the continuous spatial distribution pattern of the specific elevated facilities in each road section.

[0021] By integrating the continuous spatial distribution patterns of the specific elevated facilities in each road section and the corresponding appearance inspection results, potential problems can be identified.

[0022] The further technical solution is as follows: the continuous spatial distribution pattern of the specific elevated facilities in each section includes the detection results of structural defects such as missing, loose or abnormal installation; the appearance inspection results include the detection results of cracks, peeling, corrosion, fading or physical deformation.

[0023] The further technical solution is as follows: The specific elevated facility is analyzed for appearance using the different road section facility material types and standard appearance characteristics recorded in the individual appearance benchmark model to obtain appearance inspection results, including:

[0024] By using the different road section facility material types and standard appearance features recorded in the individual appearance benchmark model, the appearance data of the specific elevated facility is compared with the preset standard appearance features to detect and locate appearance damage that exceeds the normal fluctuation range, so as to obtain the appearance inspection results.

[0025] The further technical solution is as follows: generating an anomaly report based on the potential problem includes:

[0026] The basic situation is formed by combining the continuous spatial distribution pattern of the specific elevated facilities in each road section with the corresponding appearance inspection results;

[0027] The potential problems are created as problem items and combined with the basic information to form an anomaly report. The anomaly report is stored, wherein the problem item includes at least one of the following: location information, associated image evidence, anomaly type, and severity level.

[0028] The further technical solution is as follows: when storing the anomaly report, the anomaly report is stored in the digital asset file of the corresponding specific elevated facility.

[0029] The further technical solution is as follows: comparing the inspection data with a preset elevated facility geographic information database to identify and determine specific elevated facilities and components, and to determine the target inspection area, including:

[0030] The GPS location data in the inspection data is matched with a preset elevated facility geographic information database to determine the specific elevated facility and station number corresponding to the image in the video stream.

[0031] The image is analyzed using a target recognition model, and all key components, including the road surface and guardrails, are labeled to generate detection results with semantic labels.

[0032] The road area is determined based on the location of the inspection vehicle, the guardrail boundaries are defined on both sides, and the detection distance is adaptively set based on the depth of field and image resolution to clarify the target range of the inspection.

[0033] This invention also provides an anomaly detection system for elevated facilities based on environmental feature fusion, comprising:

[0034] The acquisition unit is used to acquire video streams, GPS location data, and timestamp information collected by equipment on the dynamic inspection vehicle to obtain inspection data.

[0035] The matching unit is used to compare the inspection data with a preset elevated facility geographic information database to identify and determine specific elevated facilities and components, and to determine the target range for inspection.

[0036] The detection unit is used to perform dual-channel detection on the specific elevated facilities based on an individual appearance benchmark model and an individual spatial relationship model, in order to identify potential problems. This includes: comparing the material types and standard appearance characteristics of the specific elevated facilities in each road segment based on the individual appearance benchmark model to obtain appearance detection results; tracking and recording the continuous spatial distribution patterns of the specific elevated facilities in each road segment based on the individual spatial relationship model; and integrating the continuous spatial distribution patterns of the specific elevated facilities in each road segment with the corresponding appearance detection results to identify potential problems.

[0037] A generation unit is used to generate an anomaly report based on the potential problem.

[0038] The advantages of this invention compared to existing technologies are as follows: By integrating video streams, GPS locations, and timestamp data collected by dynamic inspection vehicles and comparing them with a pre-set geographic information database of elevated facilities, this invention accurately identifies and locates specific elevated facilities and their components, and determines the inspection target range. It employs a dual-channel detection method using a dynamically learned individual appearance benchmark model and a spatial relationship model. On the one hand, it ensures consistency in appearance by comparing material types and standard appearance features; on the other hand, it tracks and records continuous spatial distribution patterns to capture layout changes. The combination of these two methods effectively identifies potential problems. Finally, a detailed anomaly report is generated based on these problems. This method significantly improves the accuracy and adaptability of elevated facility inspection, providing solid and reliable data support and maintenance assurance for the digital management of facilities, and achieving end-to-end optimization from data collection to problem identification and maintenance decision-making.

[0039] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A flowchart illustrating the method for determining anomalies in elevated facilities based on environmental feature fusion, as provided in an embodiment of the present invention.

[0042] Figure 2 A schematic diagram illustrating the working principle of establishing an individual appearance benchmark model provided in an embodiment of the present invention;

[0043] Figure 3 A schematic diagram illustrating the dynamic updating principle of the individual appearance benchmark model provided in this embodiment of the invention;

[0044] Figure 4 This is a schematic diagram of environmental monitoring for elevated streetlights provided in an embodiment of the present invention;

[0045] Figure 5 This is a schematic diagram illustrating the absence of facilities as provided in an embodiment of the present invention;

[0046] Figure 6 This is a schematic block diagram of an elevated facility anomaly detection system based on environmental feature fusion provided in an embodiment of the present invention.

[0047] Figure 7 A schematic block diagram of a computer device provided for an embodiment of the present invention. Detailed Implementation

[0048] 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, not all, of the embodiments of the present invention. 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.

[0049] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0050] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0051] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0052] Please see Figure 1 , Figure 1 This is a flowchart illustrating the elevated facility anomaly detection method based on environmental feature fusion provided in this invention. By integrating video streams, GPS location data, and timestamp information collected by dynamic inspection vehicles, combined with a pre-set geographic information database, precise identification and location of specific elevated facilities and their components are achieved. Utilizing a dual-channel detection method based on individual appearance benchmark models and individual spatial relationship models, not only can detailed analysis of material types and standard appearance characteristics be performed, but the spatial distribution patterns of the facilities can also be tracked, thereby effectively identifying potential problems. Furthermore, this method generates detailed anomaly reports, including key information such as location information, image evidence, anomaly type, and severity level, providing a scientific basis for the maintenance of elevated facilities. This method significantly improves the accuracy and adaptability of elevated facility detection, provides reliable assurance for the digital management of elevated facilities, and ensures the safe and stable operation of the facilities.

[0053] Figure 1 This is a flowchart illustrating the method for determining anomalies in elevated facilities based on environmental feature fusion, provided in an embodiment of the present invention. Figure 1 As shown, the method includes the following steps S110 to S140.

[0054] S110. Obtain video streams, GPS location data, and timestamp information collected by equipment on the dynamic inspection vehicle to obtain inspection data.

[0055] In this embodiment, the inspection data refers to a comprehensive dataset consisting of video streams, GPS location data, and timestamp information collected by equipment on the dynamic inspection vehicle. Specifically:

[0056] Video stream: This is a continuous sequence of images captured in real time by visual sensors such as cameras mounted on inspection vehicles, covering various angles and details of the elevated facilities. This video data provides an intuitive data source for subsequent target identification and environmental feature extraction, and is the foundation for analyzing the appearance damage and structural anomalies of the elevated facilities.

[0057] GPS location data: The Global Positioning System (GPS) devices equipped on the inspection vehicles provide precise location coordinates, ensuring that each video stream is matched with a specific geographical location. This step is crucial because it not only helps identify the specific elevated structures and their components shown in the video streams but also provides a geographic reference framework for subsequent data processing and analysis.

[0058] Timestamp Information: To ensure that all collected data can be accurately mapped to specific points in time, inspection equipment automatically adds timestamps. This is crucial for tracking the changing trends of elevated facilities over time, recording maintenance history, and providing a time-based basis for generating anomaly reports.

[0059] In summary, by integrating video streams, GPS location data, and timestamp information, the inspection data enables comprehensive, multi-angle monitoring of elevated facilities, laying the foundation for constructing an inspection method based on "individual appearance benchmarks" and "individual spatial relationship benchmarks." This method not only significantly improves inspection accuracy but also enhances system adaptability and expands inspection dimensions, thereby effectively improving the level and efficiency of digital management of elevated facilities.

[0060] S120. The inspection data is compared with the preset elevated facility geographic information database to identify and determine the specific elevated facilities and components, and to determine the target range of the inspection.

[0061] In this embodiment, the elevated infrastructure geographic information database refers to a database containing detailed geographic information and attribute data of all known elevated roads and their components. This database includes not only the location coordinates of the elevated bridges (such as GPS coordinates), but also the specific structural information of each elevated bridge, such as its length, width, height, design standards, material type, and the specific location and specifications of various facilities (such as road surface, expansion joints, curbs, guardrails, sound barriers, etc.). This information is crucial for achieving accurate matching of inspection data.

[0062] In one embodiment, step S120 described above may include steps S121 to S123.

[0063] S121. Match the GPS location data in the inspection data with the preset elevated facility geographic information database to determine the specific elevated facility and station number corresponding to the image in the video stream.

[0064] First, the GPS location data in the video stream uploaded by the dynamic inspection vehicle is compared with a pre-set geographic information database of elevated facilities to determine the specific elevated facility and its exact station number (e.g., K10+000~K10+500) corresponding to the video stream. This step ensures that subsequent analysis is performed on the correct section of the elevated facility by accurately matching the geographic location, thus providing a foundation for accurate identification.

[0065] S122. Analyze the image using a target recognition model and label all key components, including the road surface and guardrail, to generate detection results with semantic labels.

[0066] In this embodiment, the image confirmed in step S121 is then further analyzed using a target recognition model. This model identifies each element in the image frame by frame and labels key components such as road surface, expansion joints, curbs, and guardrails, generating detection results with semantic tags.

[0067] In this embodiment, the detection result refers to the processed image, which contains the precise location and status description of each elevated facility component, such as whether there is damage or abnormality. This result with semantic information facilitates subsequent data analysis and decision-making.

[0068] The detection results clearly identify the specific locations and current states of key components (such as road surfaces and guardrails) in video images. This is crucial for identifying any damage or anomalies. By comparing detection results at different time points, individual appearance benchmark models and individual spatial relationship models can be constructed and maintained for specific road sections and facility types. This helps identify material types, standard appearance features, and spatial relationships between facilities, thus providing a foundation for subsequent comparisons and anomaly detection. The detection results form the basis for environmental consistency testing and appearance damage detection. By comparing real-time data with preset benchmarks, the system can accurately identify structural defects and localized damage, and make corresponding handling decisions accordingly. The detection results are also used to generate basic situation reports and anomaly detail reports. These reports include the distribution of elevated facilities, an overview of the appearance materials of major facilities, and all detected anomalies. Each anomaly is accompanied by detailed location information, related image evidence, anomaly type, and severity level. This not only benefits subsequent maintenance and repair work but also provides important data for the digital asset management of elevated facilities.

[0069] S123. Determine the road area based on the location of the inspection vehicle, define the guardrail boundaries on both sides, and adaptively set the detection distance based on depth of field and image resolution to clarify the target range of the inspection.

[0070] Finally, based on the relative position of the inspection vehicle in the image, the road area shown in the current video frame is determined, and the guardrail boundaries are defined to both sides. Combining depth information and image resolution, the detection distance is adaptively set to clearly define the target area for this inspection. This ensures that the detection process focuses on the actual areas requiring attention, avoids unnecessary waste of computational resources, and improves detection efficiency and accuracy.

[0071] In summary, step S120, by comparing the inspection data with the elevated facility geographic information database, achieves precise location and identification of specific elevated facilities and their components, thereby clarifying the target scope of the inspection. This process forms the basis for building an individualized benchmark database and performing dual-channel anomaly detection, and is of great significance for improving detection accuracy and system adaptability.

[0072] S130. Based on a dynamic learning-based individual appearance benchmark model and an individual spatial relationship model, dual-channel detection is performed on the specific elevated facilities to identify potential problems; including: using the dynamic learning-based individual appearance benchmark model to compare the material types and standard appearance characteristics of the specific elevated facilities in each road segment to obtain appearance detection results; using the dynamic learning-based individual spatial relationship model to track and record the continuous spatial distribution pattern of the specific elevated facilities in each road segment; and integrating the continuous spatial distribution pattern of the specific elevated facilities in each road segment with the corresponding appearance detection results to identify potential problems.

[0073] In this embodiment, potential problems refer to cosmetic damage and structural defects. Cosmetic damage includes cracks, peeling, corrosion, fading, or physical deformation, which can affect the safety and aesthetics of the facility. Structural defects include missing parts, looseness, or abnormal installation, which can lead to structural instability and affect the overall service life and safety of the facility.

[0074] The continuous spatial distribution pattern of specific elevated facilities within each road section refers to the spatial arrangement order of the facilities and their geometric parameters such as distance and angle, as well as their consistency over time. Examples include the standard spacing between guardrails and the linear arrangement pattern of streetlights.

[0075] The corresponding appearance inspection results refer to the evaluation results on the surface condition of the facility obtained by comparing real-time collected appearance data with preset standard appearance characteristics. This includes, but is not limited to, the presence or absence of cracks, the degree of material peeling, and the severity of corrosion.

[0076] The continuous spatial distribution pattern of the specific elevated facilities in each section includes the results of the detection of structural defects such as missing, loose, or abnormal installation; the results of the appearance inspection include the results of the detection of cracks, peeling, corrosion, fading, or physical deformation.

[0077] In one embodiment, step S130 described above may include steps S131 to S133.

[0078] S131. Using the different road section facility material types and standard appearance characteristics recorded in the individual appearance benchmark model, perform appearance analysis on the specific elevated facility to obtain appearance inspection results.

[0079] Specifically, by using the different road section facility material types and standard appearance features recorded in the individual appearance benchmark model, the appearance data of the specific elevated facility is compared with the preset standard appearance features to detect and locate appearance damage that exceeds the normal fluctuation range, so as to obtain the appearance inspection results.

[0080] The specific operations at this stage involve extracting images from the video stream captured by the inspection equipment and determining the specific road segment based on GPS positioning information. Then, the appearance features in these images are compared with standard appearance features stored in the individual appearance benchmark model. For example, the surface condition of road surfaces, guardrails, and other structural components is inspected to detect and locate appearance damage exceeding normal fluctuations, such as cracks, peeling, corrosion, fading, or physical deformation. Once an anomaly is detected, the system generates a detailed appearance inspection report, which includes not only all discovered appearance damage and its location (station / GPS), but also the severity level.

[0081] The continuous spatial distribution pattern of the specific elevated facilities in each section includes the results of the detection of structural defects such as missing, loose, or abnormal installation; the appearance inspection results include the results of the detection of cracks, peeling, corrosion, fading, or physical deformation.

[0082] S132. Monitor the spatial distribution pattern and structural characteristics of the specific elevated facilities in real time, and compare them with the data in the individual spatial relationship model to identify abnormal changes or deviations that affect the structural integrity of the facilities, so as to obtain the continuous spatial distribution pattern of the specific elevated facilities in each road section.

[0083] During this process, the system needs to monitor the spatial distribution patterns of specific elevated facilities within each road section in real time. This includes, but is not limited to, the spacing, relative positions, and linear arrangement patterns between facilities. By comparing the observed spatial relationships with standard data in the individual spatial relationship model, any significant deviations can be effectively identified. If missing, loose, or abnormal installation conditions are found, they are identified as structural defects, and the corresponding detection results are recorded. This method not only detects problems promptly but also ensures the accuracy of the detection.

[0084] S133. Integrate the continuous spatial distribution pattern of the specific elevated facilities in each road section and the corresponding appearance inspection results to identify potential problems.

[0085] By integrating the results from both aspects of the inspection, a comprehensive assessment of the overall health of the specific elevated facility is conducted, identifying all potential problem areas. At this stage, the system combines visual inspection results (such as cracks, peeling, corrosion, etc.) and spatial distribution pattern inspection results (such as missing parts, looseness, installation anomalies, etc.) to generate a complete inspection report. This method accurately identifies all potential problems and their locations (station number / GPS), along with relevant image evidence and severity levels. The final inspection results are archived in the elevated facility's digital asset file for subsequent maintenance and management.

[0086] The entire process not only improves the accuracy of detection and the adaptability of the system, but also achieves integrated intelligent diagnosis of structural anomalies and local damage through dual-channel detection of "environmental consistency" and "appearance damage." This method ensures that elevated facilities can maintain good operating conditions over a long period of time, while also providing solid data support and technical assurance for future maintenance work.

[0087] In one embodiment, the process of establishing the above-mentioned individual appearance benchmark model includes:

[0088] Acquire multiple frames of images;

[0089] Analyze multiple consecutive frames of images to identify and record the specific material types and standard appearance characteristics of various facilities in different road sections in order to obtain feature data;

[0090] Based on the aforementioned feature data, an individualized appearance benchmark library is constructed for each road segment and its internal different facility types. When there are changes in the appearance of facilities or updates to materials, the individualized appearance benchmark library is updated.

[0091] In this embodiment, feature data refers to all relevant information about the appearance of a facility extracted from multiple frames of images, including but not limited to material type, color, surface condition, and texture. This data forms the basis for building an individualized appearance benchmark library for subsequent comparison and anomaly detection. Specifically, firstly, the system extracts consecutive multi-frame images from the video stream collected by the inspection equipment. These images are taken at different times and under different environmental conditions to ensure that the appearance characteristics of the facility are captured in various states. Next, these consecutive multi-frame images are analyzed in depth to identify and record the specific material types and standard appearance characteristics of various facilities in different road sections. This step involves using computer vision technology to detect and classify facilities in the images, such as road surfaces, guardrails, and sound barriers, and to record their surface condition, color, texture, and other details. Based on the feature data obtained in the previous step, an individualized appearance benchmark library is built for each road section and the different facility types within it. This means that for each specific elevated facility or road section, there will be a dedicated database containing the standard appearance characteristics of that facility or road section. When there are changes to the appearance of the facility (such as color changes due to maintenance) or material updates, the system will automatically update the individualized appearance benchmark library to keep the data up-to-date and accurate.

[0092] Please see Figures 2 to 3 For example, regarding soundproof panels, the system learned from existing elevated structures on specific road sections, particularly the "transparent-sound-absorbing composite soundproof panel." Through this process, the system extracted various characteristics of the structure, such as material and color, and based on this information, created a unique individual appearance benchmark model for that road section. This individual appearance benchmark model serves as a reference standard for subsequent inspections, ensuring accurate identification and reporting of anomalies. The image shows the individual appearance benchmark model of the soundproof panel, belonging to elevated structure ABCD; used on road sections KAB+CDE~KFG+HIJ; structure type: transparent-sound-absorbing composite soundproof panel; material and color: aluminum alloy (122,122,122), glass; standard dimensions: rectangle, 2.5m long, 2m wide; creation time: KLMN-OP-QR ST:UV:WX; update time: abcd-ef gh:ij:kl; Figure 3This further demonstrates a key capability of the system during continuous inspections: when the original "transparent-sound-absorbing composite sound insulation panel" is replaced with a "fully transparent sound insulation panel," the system can automatically detect this change and update the individualized baseline for that road segment based on the new situation. This is done to prevent false alarms caused by facility changes, ensuring the accuracy and reliability of the monitoring system. Figure 3 Two individual appearance benchmark models of sound insulation panels are presented, one of which is Figure 2 The content is consistent with the previous one, while the other content is as follows: the elevated road is A'B'C'D'; the road section used is KA'B'+C'D'E'~KF'G'+H'I'J'; the facility type is a fully transparent soundproof panel; the material and color are glass; the standard size is a rectangle with a length of 1.5m and a width of 1m; the creation time is KL'M'N'-O'P'-Q'R' S'T':U'V':W'X'; and the update time is a'b'c'd'-e'f' g'h':i'j':k'l'.

[0093] This case vividly illustrates the core mechanism of the invention: the individual appearance benchmark model is dynamically updated as facilities change. This mechanism endows the system with flexibility and adaptability, enabling it to automatically adjust its detection standards according to changes in actual facilities, achieving efficient and accurate monitoring and management without manual intervention. This not only improves monitoring efficiency but also reduces maintenance costs, demonstrating the significant advantages brought about by technological advancements. In this way, the system can maintain a high degree of sensitivity to various changes over a long period, ensuring effective monitoring of the status of public facilities.

[0094] In one embodiment, the process of establishing the above-mentioned individual spatial relationship model includes:

[0095] List all facilities with spatial continuity within the current road segment, and perform time-series tracking and trajectory fitting for each type of continuous facility to determine the standard spacing, relative position, and linear arrangement pattern of the facilities within their distribution road segment, thus generating analysis results;

[0096] Based on the analysis results, a spatial relationship benchmark database is established for different facility types and road segments, forming individual spatial relationship models. When a change in spatial relationship is detected, the corresponding spatial relationship benchmark database is updated.

[0097] In this embodiment, the analysis results refer to the standard spacing, relative positions, and linear arrangement patterns obtained by temporally tracking and fitting the trajectories of facilities with spatial continuity. This information is used to construct and update individual spatial relationship models, ensuring that the models accurately reflect the actual state of the facilities, thereby improving the system's detection accuracy and adaptability.

[0098] Specifically, the system first identifies a list of all spatially continuous facilities within the current road segment, such as guardrails, streetlights, and road signs. These facilities have certain spatial relationships, such as spacing and relative positions. For each type of continuous facility, the system uses time-series tracking and trajectory fitting techniques to determine the standard spacing, relative position, and linear arrangement pattern of these facilities within their distribution segment. This process requires processing a large amount of image data and using algorithms to calculate precise spatial relationships. Based on the above analysis results, a spatial relationship benchmark database is established by facility type and road segment, forming an individual spatial relationship model. This model records the spatial relationship information of each facility within a specific road segment. Once any spatial relationship change is detected (such as facility movement due to construction), the corresponding spatial relationship benchmark database is updated accordingly.

[0099] S140. Generate an anomaly report based on the potential problem.

[0100] In one embodiment, step S140 described above may include steps S141 to S142.

[0101] S141. The continuous spatial distribution pattern of the specific elevated facilities in each road section and the corresponding appearance inspection results are combined to form a basic situation.

[0102] This process first integrates spatial relationship models (such as standard spacing, relative positions, and linear arrangement patterns) and appearance benchmark models (such as material types and standard appearance features) of elevated facilities obtained through the "individualized benchmark library construction method." This information collectively constitutes the basic information for each section of elevated facility. For example, for a road segment within a specific chainage range, the system records the specific layout and expected appearance of facilities such as expansion joints, curbs, and guardrails within that segment. This information not only helps in understanding the overall layout of the current facilities but also provides a basis for comparison in subsequent anomaly detection.

[0103] S142. Create problem items for the potential problems and combine them with the basic information to form an anomaly report, and store the anomaly report, wherein the problem items include at least one of location information, associated image evidence, anomaly type, and severity level.

[0104] Specifically, when storing the anomaly report, the anomaly report is stored in the digital asset file of the corresponding specific elevated facility.

[0105] After identifying any deviation from the normal state, the system will automatically generate a problem item. Each problem item contains at least the following key elements:

[0106] Location information: Provides specific geographic location identifiers, such as station numbers or GPS coordinates.

[0107] Related image evidence: Relevant image data obtained directly from the inspection data synchronous acquisition module provides intuitive proof of the problem.

[0108] Anomaly type: Based on the results of the dual-channel anomaly detection model, it is clearly indicated whether it is a structural defect or local damage.

[0109] Severity level: Assess the urgency and scope of the problem to help decision-makers prioritize its handling.

[0110] These issues will then be combined with the previously compiled basic information to generate a detailed anomaly report. This report not only lists all detected problems but also provides relevant background information, ensuring the maintenance team has a comprehensive understanding of the context in which each problem occurred. Finally, all anomaly reports will be properly archived in the digital asset file of the corresponding specific elevated facility for long-term tracking, management, and historical data analysis.

[0111] This comprehensive analysis and report generation method can not only effectively improve the accuracy and efficiency of facility management, but also provide strong support for subsequent maintenance work.

[0112] Please see Figure 4 and Figure 5 , Figure 4 This demonstrates the system's process of learning the spatial layout of elevated streetlights on a specific road segment. Specifically, this includes extracting key parameters such as the spacing between streetlights and determining the baseline alignment axis. Based on this information, the system can establish a unique and dynamically learnable spatial relationship benchmark for that road segment. This benchmark details the ideal layout of the streetlights and serves as an important reference for subsequent monitoring and maintenance. The detailed content of the individual spatial relationship model for this streetlight is as follows: Elevated road: ABCD; Road segment: KAB+CDE~KFG+HIJ; Facility type: Streetlight; Spacing: 30m; Creation time: KLMN-OP-QR ST:UV:WX; Update time: abcd-ef gh:ij:kl.

[0113] exist Figure 5 The image shows how the system applies this spatial relationship benchmark for actual detection. When the system compares the streetlight spatial relationship benchmark with the latest acquired images, it can identify any changes deviating from the normal state. For example, if the absence of a streetlight causes a significant increase in the spacing between adjacent streetlights, the system can accurately detect this change and classify it as an anomaly of "facility missing." Once the anomaly is confirmed, the system immediately triggers an alarm mechanism to notify relevant personnel for handling. Figure 5 In addition to presentation Figure 4In addition to the individual spatial relationship model of the street light, the continuous spatial distribution pattern of the street light in each road segment is also presented. The specific content is as follows: the detection facility is a street light, the problem type is facility missing, the location is Kab+cde, the elevated road is fghij, and the collection time is klmn-op-qr st:uv:wx.

[0114] These two figures together demonstrate the powerful capabilities of this invention in automatically monitoring the integrity and layout accuracy of public facilities. By dynamically updating and comparing spatial relationship benchmarks, the system not only improves the efficiency of problem detection but also ensures the safety and normal use of public facilities, showcasing the important role of technology in enhancing the intelligence level of urban management. This mechanism ensures efficient monitoring and timely response even when facilities are altered.

[0115] By establishing and updating individualized inspection benchmarks for each elevated facility, a precise match between inspection standards and actual conditions is achieved. This method effectively reduces false alarm rates caused by individual differences in materials, specifications, etc., and improves the system's adaptability and accuracy. Through real-time identification of the facility composition in each section, the system can achieve synchronous evolution between the model and the actual condition of the facilities, thus ensuring long-term effectiveness. This allows the system not only to identify the current condition of the facilities but also to adapt to facility modifications and replacements, ensuring long-term accuracy and reliability.

[0116] The problem of limited detection dimensions: By introducing a dual-channel detection method combining "environmental consistency" and "appearance damage," integrated intelligent diagnosis of structural anomalies and localized damage is achieved. This multi-dimensional detection approach can more comprehensively assess the health status of facilities, improving detection accuracy and coverage.

[0117] By creating an individualized benchmark library matched to specific facilities and road sections, the problem of false alarms caused by individual differences in materials and specifications was solved, significantly improving judgment accuracy. Utilizing a dual-channel detection model, both localized external damage and structural defects in spatial arrangement were simultaneously identified, enabling a comprehensive assessment of facility health. The use of individualized benchmarks matched to specific facilities and road sections for comparison overcame false alarms caused by individual differences in materials and specifications, significantly improving judgment accuracy. The system can automatically identify the actual distribution of facilities within a road section and adapt to facility modifications and replacements through a dynamic update mechanism, possessing long-term evolution capabilities and ensuring the system's continued applicability. The dual-channel detection model can not only identify localized external damage but also analyze structural defects in spatial arrangement, achieving a comprehensive assessment of facility health and providing a more integrated monitoring solution.

[0118] In summary, this embodiment solves the problems of high false alarm rate, poor adaptability and single detection dimension in traditional technologies by constructing an individualized benchmark library and implementing a dual-channel anomaly detection method, and provides an efficient, accurate and highly adaptable health monitoring solution for elevated facilities.

[0119] The aforementioned method for identifying anomalies in elevated facilities based on environmental feature fusion integrates video streams, GPS location data, and timestamp data collected by dynamic inspection vehicles. This data is compared with a pre-defined geographic information database of elevated facilities to accurately identify and locate specific elevated facilities and their components, and to determine the inspection target area. A dual-channel detection method is employed, utilizing a pre-established and dynamically learnable individual appearance benchmark model and spatial relationship model. On one hand, it ensures consistency in appearance by comparing material types and standard appearance features; on the other hand, it tracks and records continuous spatial distribution patterns to capture layout changes. The combination of these two methods effectively identifies potential problems. Finally, a detailed anomaly report is generated based on these issues. This method significantly improves the accuracy and adaptability of elevated facility inspection, providing robust and reliable data support and maintenance assurance for the digital management of facilities, and achieving end-to-end optimization from data collection to problem identification and maintenance decision-making.

[0120] Figure 6 This is a schematic block diagram of an elevated facility anomaly judgment system 300 based on environmental feature fusion provided in an embodiment of the present invention. Figure 6 As shown, corresponding to the above-described method for determining anomalies in elevated facilities based on environmental feature fusion, the present invention also provides a system 300 for determining anomalies in elevated facilities based on environmental feature fusion. This system 300 includes a unit for executing the above-described method for determining anomalies in elevated facilities based on environmental feature fusion, and the system can be configured in a server. Specifically, please refer to... Figure 6 The elevated facility anomaly judgment system 300 based on environmental feature fusion includes an acquisition unit 301, a matching unit 302, a detection unit 303, and a generation unit 304.

[0121] The acquisition unit 301 is used to acquire video streams, GPS location data, and timestamp information collected by equipment on the dynamic inspection vehicle to obtain inspection data; the matching unit 302 is used to compare the inspection data with a preset elevated facility geographic information database to identify and determine specific elevated facilities and components, and to determine the target inspection range; the detection unit 303 is used to perform dual-channel detection on the specific elevated facilities based on a dynamically learned individual appearance benchmark model and an individual spatial relationship model to identify potential problems; including: comparing the material types and standard appearance characteristics of specific elevated facilities in each road segment based on the individual appearance benchmark model to obtain appearance inspection results; tracking and recording the continuous spatial distribution pattern of specific elevated facilities in each road segment based on the individual spatial relationship model; integrating the continuous spatial distribution pattern of specific elevated facilities in each road segment and the corresponding appearance inspection results to identify potential problems; and the generation unit 304 is used to generate an anomaly report based on the potential problems.

[0122] In one embodiment, the system further includes an individual appearance benchmark model establishment unit, used to acquire multiple frames of images; analyze consecutive multiple frames of images to identify and record the specific material types and standard appearance features used by various facilities in different road sections to obtain feature data; based on the feature data, construct an individualized appearance benchmark library for each road section and its different facility types, and update the individualized appearance benchmark library when there are changes in the appearance of facilities or material updates.

[0123] In one embodiment, the system further includes an individual spatial relationship model establishment unit, which is used to list all facilities with spatial continuity in the current road segment, and perform time-series tracking and trajectory fitting for each type of continuous facility to determine the standard spacing, relative position and linear arrangement pattern of the facility in its distribution road segment, and form an analysis result; establish a spatial relationship benchmark library by facility type and road segment based on the analysis result, form an individual spatial relationship model, and update the corresponding spatial relationship benchmark library when a change in spatial relationship is detected.

[0124] In one embodiment, the detection unit 303 includes:

[0125] The appearance analysis subunit is used to perform appearance analysis on the specific elevated facility using the different road segment facility material types and standard appearance characteristics recorded in the individual appearance benchmark model, so as to obtain appearance inspection results; the spatial analysis subunit is used to monitor the spatial distribution pattern and structural characteristics of the specific elevated facility in real time, and compare them with the data in the individual spatial relationship model to identify abnormal changes or deviations affecting the structural integrity of the facility, so as to obtain the continuous spatial distribution pattern of the specific elevated facility in each road segment; the integration subunit is used to integrate the continuous spatial distribution pattern of the specific elevated facility in each road segment and the corresponding appearance inspection results to identify potential problems.

[0126] In one embodiment, the external light analysis subunit is used to compare the appearance data of the specific elevated facility with the preset standard appearance features by using the different road section facility material types and standard appearance features recorded in the individual appearance benchmark model, to detect and locate appearance damage that exceeds the normal fluctuation range, so as to obtain the appearance inspection result.

[0127] In one embodiment, the generation unit 304 is used to combine the continuous spatial distribution pattern of the specific elevated facility in each road segment with the corresponding appearance inspection results to form a basic situation; to create problem items for the potential problems, and combine them with the basic situation to form an anomaly report; and to store the anomaly report, wherein the problem item includes at least one of location information, associated image evidence, anomaly type, and severity level. When storing the anomaly report, the anomaly report is stored in the digital asset file of the corresponding specific elevated facility.

[0128] In one embodiment, the matching unit 302 includes:

[0129] The information matching subunit is used to match the GPS location data in the inspection data with a preset elevated facility geographic information database to determine the specific elevated facility and station number corresponding to the image in the video stream; the analysis subunit is used to analyze the image using a target recognition model and mark all key components including the road surface and guardrails, generating detection results with semantic tags; the range determination subunit is used to determine the road surface area based on the location of the inspection vehicle, define the guardrail boundaries to both sides, and adaptively set the detection distance based on depth of field and image resolution to clarify the target range of the inspection.

[0130] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned elevated facility anomaly judgment system 300 based on environmental feature fusion and each unit can be referred to the corresponding description in the foregoing method embodiments. For the sake of convenience and brevity, it will not be repeated here.

[0131] The aforementioned elevated facility anomaly detection system 300 based on environmental feature fusion can be implemented as a computer program, which can, for example... Figure 7 It runs on the computer device shown.

[0132] Please see Figure 7 , Figure 7 This is a schematic block diagram of a computer device provided in an embodiment of this application. The computer device 500 can be a server, wherein the server can be a standalone server or a server cluster composed of multiple servers.

[0133] See Figure 7 The computer device 500 includes a processor 502, a memory, and a network interface 505 connected via a system bus 501. The memory may include a non-volatile storage medium 503 and internal memory 504.

[0134] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions that, when executed, cause the processor 502 to perform an elevated facility anomaly judgment method based on environmental feature fusion.

[0135] The processor 502 provides computing and control capabilities to support the operation of the entire computer device 500.

[0136] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute an elevated facility anomaly judgment method based on environmental feature fusion.

[0137] This network interface 505 is used for network communication with other devices. Those skilled in the art will understand that... Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device 500 to which the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0138] The processor 502 is used to run a computer program 5032 stored in the memory to implement all the steps of the elevated facility anomaly judgment method based on environmental feature fusion.

[0139] It should be understood that in the embodiments of this application, the processor 502 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0140] It will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0141] Therefore, the present invention also provides a storage medium. This storage medium can be a computer-readable storage medium. The storage medium stores a computer program, wherein when executed by a processor, the computer program causes the processor to perform all the steps of the elevated facility anomaly determination method based on environmental feature fusion.

[0142] The storage medium can be any computer-readable storage medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0143] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0144] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of each unit is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0145] The steps in the method of this invention can be adjusted, merged, or reduced in order according to actual needs. The units in the system of this invention can be merged, divided, or reduced according to actual needs. Furthermore, the functional units in the various embodiments of this invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0146] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or 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, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0147] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for judging anomalies in elevated facilities based on environmental feature fusion, characterized in that, include: The inspection data is obtained by acquiring video streams, GPS location data, and timestamp information collected by equipment on the dynamic inspection vehicle. The inspection data is compared with a pre-set geographic information database of elevated facilities to identify and determine specific elevated facilities and components, and to determine the target range for inspection. Based on a dynamic learning-based individual appearance benchmark model and an individual spatial relationship model, dual-channel detection is performed on the specific elevated facilities to identify potential problems. This includes: comparing the material types and standard appearance characteristics of the specific elevated facilities in each road segment using the dynamic learning-based individual appearance benchmark model to obtain appearance detection results; tracking and recording the continuous spatial distribution patterns of the specific elevated facilities in each road segment using the dynamic learning-based individual spatial relationship model; and integrating the continuous spatial distribution patterns of the specific elevated facilities in each road segment with the corresponding appearance detection results to identify potential problems. Generate an anomaly report based on the potential issues; The process of establishing the individual appearance benchmark model includes: Acquire multiple frames of images; Analyze multiple consecutive frames of images to identify and record the specific material types and standard appearance characteristics of various facilities in different road sections in order to obtain feature data; Based on the feature data, an individualized appearance benchmark library is constructed for each road segment and different facility types within it, and the individualized appearance benchmark library is updated when there are changes in the appearance of facilities or material updates. The process of establishing the individual spatial relationship model includes: List all facilities with spatial continuity within the current road segment, and perform time-series tracking and trajectory fitting for each type of continuous facility to determine the standard spacing, relative position, and linear arrangement pattern of the facilities within their distribution road segment, thus generating analysis results; Based on the analysis results, a spatial relationship benchmark database is established for different facility types and road segments, forming individual spatial relationship models. When a change in spatial relationship is detected, the corresponding spatial relationship benchmark database is updated.

2. The method for judging anomalies in elevated facilities based on environmental feature fusion according to claim 1, characterized in that, The dynamic learning-based individual appearance benchmark model compares the material types and standard appearance characteristics of specific elevated facilities in each road segment to obtain appearance inspection results; the dynamic learning-based individual spatial relationship model tracks and records the continuous spatial distribution patterns of specific elevated facilities in each road segment; and integrates the continuous spatial distribution patterns of the specific elevated facilities in each road segment with the corresponding appearance inspection results to identify potential problems, including: Using the different road section facility material types and standard appearance characteristics recorded in the individual appearance benchmark model, the appearance of the specific elevated facility is analyzed to obtain the appearance inspection results; The spatial distribution pattern and structural characteristics of the specific elevated facilities are monitored in real time and compared with the data in the individual spatial relationship model to identify abnormal changes or deviations that affect the structural integrity of the facilities, so as to obtain the continuous spatial distribution pattern of the specific elevated facilities in each road section. By integrating the continuous spatial distribution patterns of the specific elevated facilities in each road section and the corresponding appearance inspection results, potential problems can be identified.

3. The method for judging anomalies in elevated facilities based on environmental feature fusion according to claim 1, characterized in that, The continuous spatial distribution pattern of the specific elevated facilities in each section includes the results of the detection of structural defects such as missing, loose, or abnormal installation; the results of the appearance inspection include the results of the detection of cracks, peeling, corrosion, fading, or physical deformation.

4. The method for judging anomalies in elevated facilities based on environmental feature fusion according to claim 2, characterized in that, The method of using the different road section facility material types and standard appearance characteristics recorded in the individual appearance benchmark model to perform appearance analysis on the specific elevated facility to obtain appearance inspection results includes: By using the different road section facility material types and standard appearance features recorded in the individual appearance benchmark model, the appearance data of the specific elevated facility is compared with the preset standard appearance features to detect and locate appearance damage that exceeds the normal fluctuation range, so as to obtain the appearance inspection results.

5. The method for judging anomalies in elevated facilities based on environmental feature fusion according to claim 1, characterized in that, The step of generating an anomaly report based on the potential problem includes: The basic situation is formed by combining the continuous spatial distribution pattern of the specific elevated facilities in each road section with the corresponding appearance inspection results; The potential problems are created as problem items and combined with the basic information to form an anomaly report. The anomaly report is stored, wherein the problem item includes at least one of the following: location information, associated image evidence, anomaly type, and severity level.

6. The method for judging anomalies in elevated facilities based on environmental feature fusion according to claim 5, characterized in that, When storing the anomaly report, the anomaly report is stored in the digital asset file of the corresponding specific elevated facility.

7. The method for judging anomalies in elevated facilities based on environmental feature fusion according to claim 1, characterized in that, The step of comparing the inspection data with a pre-set geographic information database of elevated facilities to identify and determine specific elevated facilities and components, and to determine the target area for inspection, includes: The GPS location data in the inspection data is matched with a preset elevated facility geographic information database to determine the specific elevated facility and station number corresponding to the image in the video stream. The image is analyzed using a target recognition model, and all key components, including the road surface and guardrails, are labeled to generate detection results with semantic labels. The road area is determined based on the location of the inspection vehicle, the guardrail boundaries are defined on both sides, and the detection distance is adaptively set based on the depth of field and image resolution to clarify the target range of the inspection.

8. An anomaly detection system for elevated facilities based on environmental feature fusion, characterized in that, The system uses the elevated facility anomaly detection method based on environmental feature fusion as described in any one of claims 1 to 7, including: The acquisition unit is used to acquire video streams, GPS location data, and timestamp information collected by equipment on the dynamic inspection vehicle to obtain inspection data. The matching unit is used to compare the inspection data with a preset elevated facility geographic information database to identify and determine specific elevated facilities and components, and to determine the target range for inspection. The detection unit is used to perform dual-channel detection on the specific elevated facilities based on an individual appearance benchmark model and an individual spatial relationship model, in order to identify potential problems. This includes: comparing the material types and standard appearance characteristics of the specific elevated facilities in each road segment based on the individual appearance benchmark model to obtain appearance detection results; tracking and recording the continuous spatial distribution patterns of the specific elevated facilities in each road segment based on the individual spatial relationship model; and integrating the continuous spatial distribution patterns of the specific elevated facilities in each road segment with the corresponding appearance detection results to identify potential problems. A generation unit is used to generate an anomaly report based on the potential problem.

Images