Road disease data induction processing method based on road network segmentation

Abstract

The invention discloses a road network segmentation-based road disease data induction processing method. The method comprises the following steps of S1, preprocessing a system road network; step S2, loading the preprocessed data; s3, analyzing the data packet; s4, marking the data packet as a processing state; S5, identifying road information to which the acquisition point in the data packet belongs; s6, judging whether the disease belongs to a system effective disease or not; step S7, performing de-duplication or updating processing on the disease points; step S8, uniformly performing coordinate system conversion on the disease points; step S9, binding the stake number information to the disease point; step S10, processing the disease pictures; and step S11, warehousing and storing batch processing disease data. According to the method, the road network matching strategy based on the latitude and longitude of the collection point is added in the data cleaning stage, and the screening strategy for treating the disease type is combined, so that the problem that the collection point source data is directly placed in a business system and possibly separated from the road network is solved.

Description

Technical Field

[0001] The invention belongs to the technical field of road data processing, and specifically relates to a road damage data summarization and processing method based on road network segmentation. Background Art

[0002] In the business of intelligent identification and treatment of road defects, the binding of the defect data collected by the terminal with the system road network is a critical processing step. Only by accurately binding the two can the defect points, defect aggregation and other information on each road be accurately displayed in the business system. However, due to the differences in accuracy of different identification terminals, coordinate drift may occur, different coordinate systems, or invalid collection of non-specified sections. If the collected defect information is directly placed in the road network, the points will be separated from the road network, which will greatly affect the subsequent business processing. Therefore, a method for summarizing and processing road defect source data is needed to accurately bind the defect information to the business road network.

[0003] With the rapid growth of my country's total road mileage, road maintenance has become increasingly important. Faced with the diversity of road diseases, the causes are complex and varied, covering natural and human factors, such as climate change, traffic load, material aging, etc., the traditional manual detection method faces low efficiency, and detection on roads with high traffic volume may hinder traffic flow and affect road use. How to timely detect and deal with road diseases is of great significance to extending the service life of roads and reducing maintenance costs. With the development of artificial intelligence, machine learning, image processing and other technologies, the application of intelligent recognition technology in the field of road disease detection has become possible and has shown great potential and advantages.

[0004] In the prior art, due to the different collection accuracy of different data collection terminals, the longitude and latitude of the collection point may drift, or the inspection vehicle may enter the wrong route, resulting in some invalid disease data entering the road network and causing the point to be out of the road network. When summarizing the disease data into the database, the disease information is not filtered for validity, and it is directly placed on the system road network based on the longitude and latitude coordinates of the collection point and applied to subsequent business. Summary of the invention

[0005] The purpose of the present invention is to provide a method for summarizing and processing road defect data based on road network segmentation, so as to solve the problems in the prior art raised in the background technology, such as the accuracy of different data acquisition terminals. The longitude and latitude of the collection points may drift, or the inspection vehicle route may enter by mistake, resulting in some invalid defect data entering the road network, causing the point to be out of the road network, affecting the subsequent business.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: A road disease data summarization and processing method based on road network segmentation comprises the following steps: Step S1, preprocessing of the system road network; segmenting each road according to a certain distance and saving it as a json file containing a set of center coordinates and edge coordinates; Step S2, loading the pre-processed road network data and the disease type data to be identified and treated for use in the subsequent disease cleaning process; Step S3, taking out a data packet from the disease collection pool, parsing the data packet to obtain the information data in the data packet; judging whether there is a record picture set in the data packet, if so, proceeding to step S4, if not, judging it as an invalid data packet, and processing the next data packet; Step S4, marking the data packet as being processed to prevent other processing threads from repeatedly processing the data packet; Step S5, identifying the road information to which the collection point in the data packet belongs according to the road data preprocessed in step S1; Step S6, for the collection points that have been matched to the road network, judging whether they belong to effective system diseases according to the disease level calculated based on the additional disease type and the identified disease area; Step S7, deduplication or update of the defect points to avoid duplication of defect information in subsequent business road networks; Step S8, uniformly transforming the coordinate system of the diseased points; Step S9, binding the pile number information to the damaged point through the road pile number information attached to the small road section; and obtaining the structured address information of the damaged point through the API for subsequent business display; Step S10, processing the disease image, using the recorded disease identification parameters to mark the disease area frame lines in the form of single frame and multiple frames, and distinguish them with different marks, so that the disease can be displayed in a more intuitive way; Step S11, batch processing disease data is stored in the database and the data package is updated to a processed state.

[0007] According to the above technical solution, in step S3, the information data in the data packet includes collection records, record picture sets and disease identification information data.

[0008] According to the above technical solution, in step S5, the identification results of the road information to which the collection point belongs include: unidentified, indicating that this point position does not match any system road information and will not be processed; nearest point, indicating that this point position does not directly hit any road, but exists within the edge range allowed by a certain road. At this time, the coordinates of the identified point are updated to the coordinates of the nearest point to ensure that it can be attached to the system road network; within the road, indicating that this point position falls within a certain system road.

[0009] According to the above technical solution, in the road information to which the collection point belongs, the determination of unrecognized is specifically as follows: First, we conduct a preliminary screening by judging whether the distance between the coordinates of the collection point and the coordinates of the center points of all small road sections under the road is within the maximum range; only if it is within the range, we continue to make further judgments; if no small road section is matched in the end, it means that this point is not matched to any road network road, and the unrecognized mark is returned; According to the above technical solution, in the road information to which the collection point belongs, the determination within the road is specifically as follows: After the initial screening, the ray method is further used to determine whether the collection point is within the small road section. If the detection point is determined to be located inside the small road section, it means that the road to which this small road section belongs is hit, and the road markings and related road information are returned.

[0010] According to the above technical solution, the nearest point in the road information to which the collection point belongs is determined as follows: If the collection point is not within the small road segment as determined in the previous step, the closest distance and coordinates between the collection point and the fence coordinates of this small road segment are recorded, and then the remaining small road segments are judged one by one until the fence point coordinates closest to the collection point in all small road segments are found. At this time, the nearest point identifier, the coordinates of the nearest point, and the road information to which it belongs are returned.

[0011] According to the above technical solution, in step S6, the specific steps of determining whether it is a system effective disease based on the additional disease type and the disease level calculated by the identified disease area include: By comparing the disease area identified at the collection point with the system disease level configuration, it is determined whether the disease area has reached the threshold of the corresponding disease, and the disease is rated at the same time; if it has not reached the level of disease control, the processing is skipped, and if it has reached the level of disease control, the processing continues.

[0012] According to the above technical solution, in step S7, the deduplication or update processing of the diseased points is specifically as follows: First, deduplication is performed within this batch of data sets. The specific deduplication method is as follows: a unique identification mark is generated by concatenating the road name + section name + small road section + disease type, and then it is put into the existing identification set of this batch for matching and searching. If it already exists, the processing is skipped. If it does not exist, the processing continues and the current identification is stored in the identification set of this batch; Secondly, deduplicate historical diseases: form a search formula using road name + section name + small road section + disease type, and put it into the system disease table for joint conditional search to determine whether the same historical disease already exists in the system. If it does, skip the process. At this point, deduplication is completed.

[0013] According to the above technical solution, in step S8, the coordinate system of the defect points is uniformly converted as follows: the conversion of the coordinate system is realized by a user-defined conversion formula, and the conversion formula includes coordinate verification, offset calculation, radian conversion, coordinate conversion formula and coordinate recovery.

[0014] According to the above technical solution, in step S10, the process of processing the disease image is specifically as follows: First, use the object conversion method of Graphics2D to convert the target image stream to facilitate subsequent processing; Traverse the original disease annotation information set corresponding to the image, perform frame line annotation processing on the data in the original disease annotation information set one by one, and calculate and set the annotation position of the rectangular frame according to the upper, lower, left and right margin positions annotated by the original disease annotation information set. The specific implementation method is: round off the X and Y coordinates of the frame selection point in the collected information to obtain the coordinates of the upper left corner of the annotation wireframe; then round off the (bottom edge - top edge) distance and (right edge - left edge) distance to obtain the height and width of the annotation wireframe; Annotation box color setting, each annotation line frame is colored with a different color; After the wireframe is initialized, wireframe rendering and annotation are performed; Upload annotated images to OBS for image storage and access; after image annotating, release system resources.

[0015] Compared with the prior art, the present invention has the following beneficial effects: In the present invention, by adding a road network matching strategy based on the longitude and latitude of the collection point in the data cleaning stage, and combining it with a screening strategy for the type of disease control, the problem of being out of the road network that may arise when the collection point source data is directly placed in the business system is solved; and the nearest point strategy during road network matching can also correct some disease points with coordinate drift to the road network, which provides better data validity support for subsequent business display and data summary statistics. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of the algorithm of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Embodiment 1 like Figure 1 As shown, a road disease data summarization and processing method based on road network segmentation includes the following steps: Step S1, preprocessing of the system road network; segmenting each road according to a certain distance and saving it as a json file containing a set of center coordinates and edge coordinates; Step S2, loading the pre-processed road network data and the disease type data to be identified and treated for use in the subsequent disease cleaning process; Step S3, taking out a data packet from the disease collection pool, parsing the data packet to obtain the information data in the data packet; judging whether there is a record picture set in the data packet, if so, proceeding to step S4, if not, judging it as an invalid data packet, and processing the next data packet; Step S4, marking the data packet as being processed to prevent other processing threads from repeatedly processing the data packet; Step S5, identifying the road information to which the collection point in the data packet belongs according to the road data preprocessed in step S1; Step S6: for the collection point of the matched road network, judging whether it belongs to the effective system disease according to the additional disease type and the disease level calculated by the identified disease area; the additional disease type is shown in Table 1; Table 1 Additional disease types Step S7, deduplication or update of the defect points to avoid duplication of defect information in subsequent business road networks; Step S8, uniformly transforming the coordinate system of the diseased points; Step S9, binding the pile number information to the damaged point through the road pile number information attached to the small road section; and obtaining the structured address information of the damaged point through the API for subsequent business display; Step S10, processing the disease image, using the recorded disease identification parameters to mark the disease area frame lines in the form of single frame and multiple frames, and distinguish them with different marks, so that the disease can be displayed in a more intuitive way; Step S11, batch processing disease data is stored in the database and the data package is updated to a processed state.

[0019] In the present invention, by adding a road network matching strategy based on the longitude and latitude of the collection point in the data cleaning stage, and combining it with a screening strategy for the type of disease control, the problem of being out of the road network that may arise when the collection point source data is directly placed in the business system is solved; and the nearest point strategy during road network matching can also correct some disease points with coordinate drift to the road network, which provides better data validity support for subsequent business display and data summary statistics.

[0020] Embodiment 2 This embodiment is a further refinement of the first embodiment.

[0021] In step S3, the information data in the data packet includes collection records, record picture sets and disease identification information data.

[0022] In step S5, the identification results of the road information to which the collection point belongs include: unidentified, indicating that this point does not match any system road information and will not be processed; nearest point, indicating that this point does not directly hit any road, but exists within the edge range allowed by a certain road. At this time, the coordinates of the identified point are updated to the coordinates of the nearest point to ensure that it can be attached to the system road network; within the road, indicating that this point falls within a certain system road.

[0023] In the road information to which the collection point belongs, the determination of unrecognized is as follows: First, we conduct a preliminary screening by judging whether the distance between the coordinates of the collection point and the coordinates of the center points of all small road sections under the road is within the maximum range; only if it is within the range, we continue to make further judgments; if no small road section is matched in the end, it means that this point is not matched to any road network road, and the unrecognized mark is returned; In the road information to which the collection point belongs, the determination within the road is specifically as follows: After the initial screening, the ray method is further used to determine whether the collection point is within the small road section. If the detection point is determined to be located inside the small road section, it means that the road to which this small road section belongs is hit, and the road markings and related road information are returned.

[0024] In the road information to which the collection point belongs, the nearest point is determined as follows: If the collection point is not within the small road segment as determined in the previous step, the closest distance and coordinates between the collection point and the fence coordinates of this small road segment are recorded, and then the remaining small road segments are judged one by one until the fence point coordinates closest to the collection point in all small road segments are found. At this time, the nearest point identifier, the coordinates of the nearest point, and the road information to which it belongs are returned.

[0025] In step S6, the specific steps of determining whether it is a system effective disease according to the disease level calculated based on the attached disease type and the identified disease area include: By comparing the disease area identified at the collection point with the system disease level configuration, it is determined whether the disease area has reached the threshold of the corresponding disease, and the disease is rated at the same time; if it has not reached the level of disease control, the processing is skipped, and if it has reached the level of disease control, the processing continues.

[0026] In step S7, the deduplication or update processing of the diseased points is specifically as follows: First, deduplication is performed within this batch of data sets. The specific deduplication method is as follows: a unique identification mark is generated by concatenating the road name + section name + small road section + disease type, and then it is put into the existing identification set of this batch for matching and searching. If it already exists, the processing is skipped. If it does not exist, the processing continues and the current identification is stored in the identification set of this batch; Secondly, deduplicate historical diseases: form a search formula using road name + section name + small road section + disease type, and put it into the system disease table for joint conditional search to determine whether the same historical disease already exists in the system. If it does, skip the process. At this point, deduplication is completed.

[0027] In step S8, coordinate system conversion of the defect points is performed uniformly. Specifically, the coordinate system conversion is realized by a user-defined conversion formula, which includes coordinate verification, offset calculation, radian conversion, coordinate conversion formula and coordinate recovery.

[0028] In step S10, the process of processing the diseased image is specifically as follows: First, use the object conversion method of Graphics2D to convert the target image stream to facilitate subsequent processing; Traverse the original disease annotation information set corresponding to the image, perform frame line annotation processing on the data in the original disease annotation information set one by one, and calculate and set the annotation position of the rectangular frame according to the upper, lower, left and right margin positions annotated by the original disease annotation information set. The specific implementation method is: round off the X and Y coordinates of the frame selection point in the collected information to obtain the coordinates of the upper left corner of the annotation wireframe; then round off the (bottom edge - top edge) distance and (right edge - left edge) distance to obtain the height and width of the annotation wireframe; Annotation box color setting, each annotation line frame is colored with a different color; After the wireframe is initialized, wireframe rendering and annotation are performed; Upload annotated images to OBS for image storage and access; after image annotating, release system resources.

[0029] Embodiment 3 The inventive concept of the present invention is: Step 1: Preprocess the system road network, divide each road into 10m / segments, and save them as json files containing center coordinates and edge coordinate sets (hereinafter referred to as small road segments); Step 2: Load the pre-processed road network data and the data on the types of damage to be identified and treated for use in the subsequent damage cleaning process.

[0030] Step 3: Take a data packet (containing N collection records) from the disease collection pool. After parsing the data packet, three layers of main data can be obtained, including collection records, record image sets, and disease identification information. Since disease images are the key content of subsequent business, first determine whether the image data set under the data packet exists, otherwise skip the invalid data packet and process the next one.

[0031] Step 4: Mark the data packet as being processed to prevent other processing threads from repeatedly processing the data packet, and then start the following loop processing process (a collection record is a possible disease point information, hereinafter referred to as a collection point).

[0032] Step 5: Combine the small road section data preprocessed in the first step to identify the road information to which the collection point belongs. There are currently three types of recognition results: Unrecognized, indicating that this point does not match any system road information and will not be processed; Nearest point: This means that this point does not directly hit any road, but exists within the allowed edge range of a road (currently ≤50m, which can be adjusted according to business design). At this time, the coordinates of the recognition point are updated to the coordinates of the nearest point to ensure that it can be attached to the system road network; Inside the road, this means that this point falls within a system road.

[0033] Step 6: For the collection points of the matched road network, based on the additional disease type and the disease level calculated from the identified disease area, comprehensively judge whether it belongs to the effective disease of the system. The specific judgment is implemented as follows: By comparing the disease area identified at the collection point with the system disease level configuration, it is determined whether the area has reached the threshold of the corresponding disease, and the disease is rated (non-disease, slight, moderate, severe). If it does not reach the level of disease control, the processing is skipped. If it reaches the level of disease control, the processing continues (hereinafter referred to as the disease point).

[0034] Step 7: Considering the fact that inspection vehicles may collect data repeatedly on the same road, it is possible that the identified disease already exists in the system disease pool. Therefore, the disease points need to be deduplicated or updated. Currently, the only rule of "10m small road section + disease type" on the same road is used to deduplicate the disease to avoid duplication and confusion of disease information in the subsequent business road network. The specific deduplication process is as follows: First, de-duplicate the data within this batch: generate a unique identification tag: road name_section name_small section_disease type, such as "Beijing Avenue_Xingmin Road_110_pothole", and then put it into the existing tag set of this batch for search. If it already exists, skip the processing. If it does not exist, continue processing and store the current tag in the tag set of this batch.

[0035] Secondly, deduplicate historical diseases: put several parameters such as road name + section name + small road section + disease type into the system disease table for joint conditional search to determine whether the same historical disease already exists in the system. If it does, skip the processing. At this point, deduplication judgment is completed.

[0036] Step 8. To meet the needs of different business designs in the system in the future, the coordinate system of the damaged points will be uniformly converted. Currently, the commonly used WGS-84 and GCJ-02 coordinate systems are supported. The conversion of the coordinate system is achieved through a custom conversion formula (coordinate verification, offset calculation, radian conversion, coordinate conversion formula, and coordinate recovery).

[0037] Step 9: Use the road pile number information attached to the small road section to synchronously bind the pile number information to the defect point; and obtain the structured address information of the defect point through the third-party API for subsequent business display.

[0038] Step 10: Processing of disease images: Using the disease identification parameters in the records, the disease images are marked with single-frame and multi-frame disease area frames, and distinguished with different colors to display the diseases in a more intuitive way. The processing process is as follows: First, use Graphics2D (a third-party graphics component package) to convert the target image stream into an object for subsequent processing; The original disease annotation information set corresponding to the image is cycled, and frame line annotation is processed one by one. According to the upper, lower, left, and right margin positions of the disease annotation, the annotation position of the rectangular frame (x coordinate of the origin, y coordinate of the origin, length of the rectangle, and width of the rectangle) is calculated and set. Marking box color setting. Multiple colors are preset here, and each marking line frame is colored with a different color (g.setColor); After the wireframe is initialized, perform wireframe rendering annotation (g.dispose()); Upload annotated images to OBS (Huawei file storage system) to store and access them; After the picture is marked, the system resources are released; Step 11: Batch process the disease data into a database and save it, and update the data package to the "processed" status, related system resources.

[0039] In order to achieve effective binding of points and road networks, the system encapsulates a unified recognition method. The implementation process includes initial distance screening, ray method, nearest point judgment, etc. By comparing the coordinates of the recognized points with the small sections of each road one by one, it finally outputs three recognition results: unrecognized, nearest point, and within the road.

[0040] 1. First, perform an initial screening by judging whether the distance between the coordinates of the collection point and the coordinates of the center points of all small road sections under the road is within the maximum range (≤50m). Only if it is within the range, further judgment will be made. If no small road section is matched in the end, it means that this point has not been matched to any road network road, and the "unrecognized" mark is returned.

[0041] 2. After passing the initial screening, the ray method is further used to determine whether the collection point is within the small road section (the ray method judgment principle is shown in the figure below). If the detection point is located exactly inside the small road section, it means that the road to which this small road section belongs is hit, and the "inside the road" mark and related road information are returned.

[0042] 3. If the collection point is not within the small road section in the previous step, record the distance and coordinates between the collection point and the fence coordinates of this small road section, and then continue to judge the remaining small road sections one by one until the fence point coordinates closest to the collection point in all small road sections are found. At this time, return the "nearest point" mark, the coordinates of the nearest point, and the road information to which it belongs.

[0043] 4. The method calling side performs the next business processing based on the road network identification results.

[0044] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0045] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A road disease data summarization and processing method based on road network segmentation, characterized by: The following steps are involved: Step S1, preprocessing of the system road network; segmenting each road according to a certain distance and saving it as a file containing a set of center coordinates and edge coordinates; Step S2, loading the pre-processed road network data and the disease type data to be identified and treated for use in the subsequent disease cleaning process; Step S3, taking out a data packet from the disease collection pool, parsing the data packet to obtain the information data in the data packet; judging whether there is a record picture set in the data packet, if so, proceeding to step S4, if not, judging it as an invalid data packet, and processing the next data packet; Step S4, marking the data packet as being processed to prevent other processing threads from repeatedly processing the data packet; Step S5, identifying the road information to which the collection point in the data packet belongs according to the road data preprocessed in step S1; Step S6, for the collection points of the matched road network, judging whether it belongs to a system effective disease according to the disease level calculated based on the attached disease type and the identified disease area; Step S7, deduplication or update of the defect points to avoid duplication of defect information in subsequent business road networks; Step S8, uniformly transforming the coordinate system of the diseased points; Step S9, binding the pile number information to the damaged point through the road pile number information attached to the small road section; and obtaining the structured address information of the damaged point through the API for subsequent business display; Step S10, processing the disease image, using the recorded disease identification parameters to mark the disease area frame lines in the form of single frame and multiple frames, and distinguish them with different marks, so that the disease can be displayed in a more intuitive way; Step S11, batch processing disease data is stored in the database and the data package is updated to a processed state.

2. The method for summarizing and processing road damage data based on road network segmentation according to claim 1 is characterized by: In step S3, the information data in the data packet includes collection records, record picture sets and disease identification information data.

3. The method for summarizing and processing road damage data based on road network segmentation according to claim 1 is characterized in that: In step S5, the identification results of the road information to which the collection point belongs include: unidentified, indicating that this point does not match any system road information and will not be processed; nearest point, indicating that this point does not directly hit any road, but exists within the edge range allowed by a certain road. At this time, the coordinates of the identified point are updated to the coordinates of the nearest point to ensure that it can be attached to the system road network; within the road, indicating that this point falls within a certain system road.

4. The method for summarizing and processing road damage data based on road network segmentation according to claim 3 is characterized by: In the road information to which the collection point belongs, the determination of unrecognized is as follows: First, an initial screening is performed by judging whether the distance between the coordinates of the collection point and the coordinates of the center points of all small road sections under the road is within the maximum range; further judgment is continued only if it is within the range; if no small road section is matched in the end, it means that this point has not been matched to any road network road, and an unrecognized mark is returned.

5. The method for summarizing and processing road damage data based on road network segmentation according to claim 4 is characterized in that: In the road information to which the collection point belongs, the determination within the road is specifically as follows: After the initial screening, the ray method is further used to determine whether the collection point is within the small road section. If the detection point is determined to be located inside the small road section, it means that the road to which this small road section belongs is hit, and the road markings and related road information are returned.

6. The method for summarizing and processing road damage data based on road network segmentation according to claim 5 is characterized in that: In the road information to which the collection point belongs, the nearest point is determined as follows: If the collection point is not within the small road segment as determined in the previous step, the closest distance and coordinates between the collection point and the fence coordinates of this small road segment are recorded, and then the remaining small road segments are judged one by one until the fence point coordinates closest to the collection point in all small road segments are found. At this time, the nearest point identifier, the coordinates of the nearest point, and the road information to which it belongs are returned.

7. The method for summarizing and processing road damage data based on road network segmentation according to claim 1 is characterized by: In step S6, the specific steps of determining whether the disease belongs to a system effective disease according to the disease level calculated based on the additional disease type and the identified disease area include: By comparing the disease area identified at the collection point with the system disease level configuration, it is determined whether the disease area has reached the threshold of the corresponding disease, and the disease is rated at the same time; if it has not reached the level of disease control, the processing is skipped, and if it has reached the level of disease control, the processing continues.

8. The method for summarizing and processing road damage data based on road network segmentation according to claim 1 is characterized by: In step S7, the deduplication or update processing of the diseased points is specifically as follows: First, deduplication is performed within this batch of data sets. The specific deduplication method is as follows: a unique identification mark is generated by concatenating the road name + section name + small road section + disease type, and then it is put into the existing identification set of this batch for matching and searching. If it already exists, the processing is skipped. If it does not exist, the processing continues and the current identification is stored in the identification set of this batch; Secondly, deduplicate historical diseases: form a search formula using road name + section name + small road section + disease type, and put it into the system disease table for joint conditional search to determine whether the same historical disease already exists in the system. If it does, skip the process. At this point, deduplication is completed.

9. The method for summarizing and processing road damage data based on road network segmentation according to claim 1, characterized in that: In step S8, coordinate system conversion of the defect points is performed uniformly. Specifically, the coordinate system conversion is realized by a user-defined conversion formula, which includes coordinate verification, offset calculation, radian conversion, coordinate conversion formula and coordinate recovery.

10. The method for summarizing and processing road damage data based on road network segmentation according to claim 1, characterized in that: In step S10, the process of processing the diseased image is specifically as follows: First, use the object conversion method of Graphics2D to convert the target image stream to facilitate subsequent processing; Traverse the original disease annotation information set corresponding to the image, perform frame line annotation processing on the data in the original disease annotation information set one by one, calculate and set the annotation position of the rectangular frame according to the upper, lower, left and right margin positions annotated by the original disease annotation information set, and implement it in the following way: obtain the coordinates of the upper left corner of the annotation wireframe by rounding off the X and Y coordinates of the frame selection point in the collected information; and then obtain the height and width of the annotation wireframe by rounding off the (bottom edge - top edge) distance and the (right edge - left edge) distance respectively; Annotation box color setting, each annotation line frame is colored with a different color; After the wireframe is initialized, wireframe rendering and annotation are performed; Upload annotated images to OBS to store and access them; After the picture is marked, the system resources are released.

Images