A general trunk highway illegal building automatic investigation method based on laser point cloud

Abstract

This invention discloses an automated method for identifying illegal structures on ordinary trunk highways based on laser point clouds. The method includes: extracting the geometric boundaries of the ordinary trunk highway based on vehicle-mounted laser point cloud scanning data to define the highway land area; calculating the relative distance between the point cloud of buildings / structures and the edge line to extract laser point clouds within the highway and its building control zone; removing gross errors and local outliers from the original point cloud data to achieve effective point cloud extraction of the entire illegal structure; clustering and segmenting the filtered non-ground point cloud using an improved DBSCAN point cloud clustering method to divide each feature point cloud into independent illegal structure objects to be identified; constructing an illegal structure knowledge base based on features and identifying the types of illegal structures; and re-identifying the illegal structure image based on ATD files and the illegal structure point cloud identification results. This invention provides strong technical support for accurate and efficient automated identification and subsequent management of illegal structures on ordinary trunk highways.

Description

Technical Field

[0001] This invention belongs to the field of management and maintenance of ordinary trunk highways and investigation and management of illegal buildings, specifically involving an automated investigation method for illegal buildings on ordinary trunk highways based on laser point clouds. Background Technology

[0002] Illegal constructions on ordinary trunk highways refer to buildings and their ancillary facilities constructed within the construction control zone of ordinary trunk highways without the permission of local governments or in violation of relevant laws and regulations. Illegal constructions on ordinary trunk highways restrict visibility and transportation efficiency, and also lead to the streetification and marketization of highways, posing a significant impact and potential danger to the traffic function of ordinary trunk highways as regional corridors. Traditional methods of investigating illegal constructions on ordinary trunk highways mainly rely on manual means, which are limited by high costs, long processing times, and low accuracy, and cannot meet the needs of large-scale investigations. Meanwhile, under the broader context of "digital transportation," various provinces have carried out numerous digitalization projects for trunk highway infrastructure, creating opportunities and mature conditions for the development of automated illegal construction investigations.

[0003] Therefore, how to leverage existing vectorized results of ordinary trunk highway assets based on digital data to conduct reasonable, efficient, and accurate automated investigations of illegal structures such as illegal buildings on the roadside and overpasses on ordinary trunk highways, and systematically uncover illegal construction issues on ordinary trunk highways, is one of the urgent problems to be solved in the field of ordinary trunk highway management, maintenance, and illegal construction investigation and management.

[0004] According to existing literature, in recent years, with the development and widespread application of mobile vehicle-mounted laser scanning systems, rapid and automated road information acquisition and 3D reconstruction of road environments have been widely used. The invention disclosed in publication number CN115439672A presents an image matching method, illegal construction detection method, terminal equipment, and storage medium based on digital models, map tiles, and UAV images. By clustering digital surface models and projecting UAV images and map tiles, it obtains building difference information, providing possibilities for image acquisition and matching, and refined 3D reconstruction of road environments.

[0005] Meanwhile, current research and applications of laser scanning systems mainly focus on areas such as automated extraction of road geometry information and road intersection recognition. Research in the field of illegal construction investigation is limited, especially regarding automated illegal construction investigation targeting ordinary trunk highways and considering their specific road characteristics. The invention disclosed in CN110322503A presents a method, system, and electronic equipment for locating illegal constructions based on laser point cloud data. This method compares two flight point cloud datasets of the measured area to retrieve differences and construct and locate suspected illegal construction points. However, this invention does not consider road characteristics, and its implementation requires data from multiple time points. Without historical data, it is impossible to directly apply existing methods for rapid investigation of existing illegal constructions. Summary of the Invention

[0006] The purpose of this invention is to propose an automated method for identifying illegal structures on ordinary trunk highways based on laser point clouds. Through technologies such as scan lines, slope algorithms, DBSCAN clustering, and grid partitioning, an automated identification approach for illegal structures on ordinary trunk highways is constructed, based on the framework of "road edge line extraction – overall point cloud filtering of illegal structures – clustering and segmentation of individual point clouds of illegal structures." Furthermore, a road edge line extraction method based on scan lines and slope algorithms, and an illegal structure identification technology based on improved DBSCAN point cloud clustering and feature construction are proposed. This overcomes the drawback of difficult parameter adjustment, optimizes clustering results, and achieves rapid segmentation of the filtered overall point cloud, providing strong technical support for the reasonable, efficient, and accurate automated identification of illegal structures on ordinary trunk highways.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0008] An automated method for detecting illegal structures on ordinary trunk highways based on laser point clouds includes the following steps:

[0009] Step A: Based on the vehicle-mounted laser point cloud scanning data, extract the geometric boundaries of ordinary trunk highways and select the highway land area.

[0010] Step B: Calculate the relative distance between the point cloud of the building / structure and the edge line, and extract the laser point cloud within the highway and highway building control area; remove gross errors and local outliers from the original point cloud data to achieve overall effective point cloud extraction of the illegal construction object;

[0011] Step C involves clustering and segmenting the filtered non-ground point cloud based on the improved DBSCAN point cloud clustering and segmenting, dividing each ground feature point cloud into independent illegal construction objects to be identified.

[0012] Step D: Construct a knowledge base of illegal constructions based on features and identify the types of illegal constructions;

[0013] Step E: Based on the ATD file and the point cloud recognition results of the illegal buildings, perform image re-recognition of the illegal buildings.

[0014] Furthermore, in step A, point cloud elevation, point cloud density, slope difference analysis, and road edge point clustering methods are used to extract the geometric boundaries of ordinary trunk highways.

[0015] Furthermore, in step A, the scope of highway land use is determined according to the highway construction control zone of "no less than 20 meters for national highways and no less than 15 meters for provincial highways".

[0016] Furthermore, in step B, gross errors and local outliers in the original point cloud data are removed by combining a pass-through filter with a radius filter.

[0017] Furthermore, in step C, the point cloud clustering and segmentation based on the improved DBSCAN includes the following steps:

[0018] Step C1, based on mesh generation:

[0019] (1) Based on the distribution characteristics of point clouds and data processing requirements, establish a one-to-one correspondence between the serial number and coordinate value of spatial points, and then set the filtered point cloud;

[0020] (2) After the index is established, the three-dimensional range is confirmed and divided into equal intervals, forming a three-dimensional grid with a 1-meter interval in the vehicle travel direction, a 0.2-meter interval in the vertical direction, and a 0.2-meter interval in the height direction.

[0021] (3) Based on the three-dimensional information of the points, match them into their respective grids to form corresponding numbers.

[0022] Step C2, DBSCAN clustering and segmentation:

[0023] (1) Based on the gridding, labels are added to each point in the point cloud according to the point density in the grid to achieve segmentation;

[0024] (2) Extract the points of each label from the overall point cloud based on the label. The point cloud composed of the points of each label will correspond to an object to be identified, which will serve as the basis for subsequent point cloud recognition.

[0025] Furthermore, in step D, the construction of a knowledge base for illegal constructions and the identification of illegal construction types based on features includes the following steps:

[0026] Step D1: For illegal roadside structures, the following procedure shall be followed for identification:

[0027] (1) Retrieve the road edge point A that is closest to the point O where the point cloud of the building / structure contacts the ground, search for the road edge point on the opposite side, retrieve the road edge point B that is closest to point A, and determine whether the building is inside or outside the road based on the coordinates of ground points A, B, and O.

[0028] (2) If it is located within the edge of the highway, it is considered an illegal structure;

[0029] (3) If it is outside the edge line of the highway, calculate and determine the minimum distance between OA and OB;

[0030] (4) If the minimum distance is less than the distance requirement of the highway building control zone, it is determined to be an illegal building; otherwise, it is determined to be a non-illegal building.

[0031] Step D2: For illegal overhead pipelines, the following procedure shall be followed for identification:

[0032] (1) Retrieve the road surface point A that is closest to the lowest point O of the point cloud of the overpass, calculate the distance between A and O, and determine whether the overpass is within or outside the highway construction clearance:

[0033] (2) If the distance between OA is less than the road clearance distance requirement, it is judged as an illegal building; otherwise, it is judged as a non-illegal building.

[0034] (3) For overhead power transmission lines, determine whether the distance between OA is less than the minimum vertical distance requirement from the road surface at the intersection of the power transmission line conductor and the highway. If it is less, it is determined to be an illegal construction; otherwise, it is determined to be a non-illegal construction.

[0035] Step D3: Based on the spatial topological relationship, physical size, and point cloud characteristics of different illegal constructions and the ground, an illegal construction knowledge base is established to further determine the types of illegal constructions. The point cloud characteristics of three typical illegal construction characteristics in the knowledge base are shown in Table 2.

[0036] Table 2. Point cloud features of three typical types of illegal constructions.

[0037]

[0038] Furthermore, in step E, based on the ATD file and the point cloud recognition results of the illegal buildings, real-scene images of the corresponding locations of the illegal buildings are extracted, and the illegal building type is identified and classified in the real-scene images through the target recognition model, thus completing the re-recognition of the illegal building images.

[0039] Beneficial effects:

[0040] This invention presents an automated method for investigating illegal constructions on ordinary trunk highways based on laser point clouds. Targeting the characteristics of laser point cloud data and the characteristics of illegal buildings on ordinary trunk highways, it constructs an automated investigation approach based on characteristic analysis. Addressing the key issues of automated investigation—road edge line extraction and illegal building point cloud clustering extraction—it proposes an improved edge line extraction technique based on scan lines and a grid-based DBSCAN point cloud clustering and segmentation technique. Furthermore, it proposes a feature-based method for constructing an illegal construction knowledge base. This invention fully utilizes existing vectorization results of ordinary trunk highway assets, systematically mines illegal construction issues on ordinary trunk highways, and provides support for the rapid and automated investigation and management of illegal constructions on ordinary trunk highways, playing a significant auxiliary role. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating the automated investigation process for illegal constructions on ordinary trunk highways based on laser point clouds, as described in an embodiment of the present invention.

[0042] Figure 2 This is a schematic diagram illustrating the extraction of road edge lines for verifying the automated investigation effect of illegal constructions on ordinary trunk highways according to an embodiment of the present invention.

[0043] Figure 3 This is a point cloud map of a road segment used to verify the effectiveness of automated screening of illegal structures on ordinary trunk highways in this embodiment of the invention.

[0044] Figure 4 This is a schematic diagram of the point cloud recognition results of an illegal building image on a section of a road that serves as a verification of the automated investigation effect of illegal buildings on ordinary trunk highways according to an embodiment of the present invention.

[0045] Figure 5 This is a schematic diagram showing the image re-identification results of illegal buildings on a section of a road that verifies the effectiveness of automated screening of illegal buildings on ordinary trunk highways according to an embodiment of the present invention. Detailed Implementation

[0046] To more clearly illustrate the purpose, technical solution, and advantages of the present invention, the specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings in the embodiments of this application.

[0047] Example 1:

[0048] See Figure 1 This invention discloses an automated method for detecting illegal constructions on ordinary trunk highways based on laser point clouds, comprising the following steps:

[0049] Step A: Based on vehicle-mounted laser point cloud scanning data, extract the geometric boundaries of ordinary trunk highways and select the highway land use area. To verify the applicability of the investigation method, an example is used on section JZ of an ordinary trunk highway in Henan Province to verify the effectiveness of automated illegal construction investigation. See the schematic diagram of the extracted road edge lines below. Figure 2 The data was obtained through bidirectional scanning using a vehicle-mounted laser measurement system. The total length of the verified road section was 137km. This section is characterized by a high concentration of villages and towns along the route, and the roadside contains a large number and variety of objects to be investigated.

[0050] Step B: Calculate the relative distance between the point cloud of buildings / structures and the edge line, extract the laser point cloud within the highway and highway construction control area, and verify the point cloud data of the road segment (see schematic diagram). Figure 3 Remove coarse-grained points and local outliers from the original point cloud data to achieve effective point cloud extraction of the entire illegal construction object.

[0051] Step C involves clustering and segmenting the filtered non-ground point cloud based on the improved DBSCAN point cloud clustering method, dividing each ground feature point cloud into independent illegal construction objects to be identified. Specific steps include:

[0052] Step C1, based on mesh generation:

[0053] (1) Based on the distribution characteristics of point clouds and data processing requirements, establish a one-to-one correspondence between the serial number and coordinate value of spatial points, and then set the filtered point cloud;

[0054] (2) After the index is established, the three-dimensional range is confirmed and divided into equal intervals, forming a three-dimensional grid with a 1-meter interval in the vehicle travel direction, a 0.2-meter interval in the vertical direction, and a 0.2-meter interval in the height direction.

[0055] (3) Based on the three-dimensional information of the points, match them into their respective grids to form corresponding numbers.

[0056] Step C2, DBSCAN clustering and segmentation:

[0057] (1) Based on the gridding, labels are added to each point in the point cloud according to the point density in the grid to achieve segmentation;

[0058] (2) Extract points from the overall point cloud based on the labels. The point cloud composed of points of each label type will correspond to an object to be identified, serving as the basis for subsequent point cloud recognition. See the schematic diagram of the point cloud recognition results for an image of an illegal building on a certain road section. Figure 4 .

[0059] Step D involves constructing a knowledge base on illegal constructions based on features and identifying the types of illegal structures. Specific steps include:

[0060] Step D1: Identify illegal roadside structures according to the following procedure:

[0061] (1) Retrieve the road edge point A that is closest to the point O where the point cloud of the building / structure contacts the ground, search for the road edge point on the opposite side, retrieve the road edge point B that is closest to point A, and determine whether the building is inside or outside the road based on the coordinates of ground points A, B, and O.

[0062] (2) If it is located within the edge of the highway, it is considered an illegal structure;

[0063] (3) If it is outside the edge line of the highway, calculate and determine the minimum distance between OA and OB;

[0064] (4) If the minimum distance is less than the distance requirement of the highway building control zone, it is determined to be an illegal building; otherwise, it is determined to be a non-illegal building.

[0065] Step D2: For illegal overhead pipelines, the following procedure shall be followed for identification:

[0066] (1) Retrieve the road surface point A that is closest to the lowest point O of the point cloud of the overpass, calculate the distance between A and O, and determine whether the overpass is within or outside the highway construction clearance:

[0067] (2) If the distance between OA is less than the road clearance distance requirement, it is judged as an illegal building; otherwise, it is judged as a non-illegal building.

[0068] (3) For overhead power transmission lines, determine whether the distance between OA is less than the minimum vertical distance requirement from the road surface at the intersection of the power transmission line conductor and the highway. If it is less, it is determined to be an illegal construction; otherwise, it is determined to be a non-illegal construction.

[0069] Step D3: Based on the spatial topological relationship, physical size, and point cloud characteristics of different illegal construction objects and the ground, establish an illegal construction knowledge base to further determine the type of illegal construction.

[0070] Step E involves re-identifying the illegal structures based on the ATD file and the illegal structure point cloud recognition results. Specifically, for the verification section, based on the ATD file and the illegal structure point cloud recognition results, real-world images of the corresponding locations of the illegal structures are extracted. Then, using a target recognition model, the real-world images are used to identify and classify the illegal structure type, as shown in Table 3. Examples of the re-identification results for illegal structure images in the verification section can be found in [link to Table 3]. Figure 5 .

[0071] Table 3. Classification and Investigation Results of Illegal Buildings in Verification Sections

[0072]

[0073] Note: Road sections not listed are those without illegal constructions.

[0074] To further verify the effectiveness of this invention, the verification road section was analyzed through manual on-site investigation, combining RTK location data. The results show that all 34 illegal constructions identified on the verification road section matched the actual situation, demonstrating that the algorithm of this invention can meet the needs of automated illegal construction investigation on ordinary trunk highways.

[0075] As can be seen from the above description of the specific embodiments in conjunction with the accompanying drawings, this invention constructs an automated screening approach for illegal buildings on ordinary trunk roads, targeting the characteristics of laser point cloud data and the characteristics of illegal buildings on ordinary trunk roads. It proposes a road edge line extraction technique for ordinary trunk roads using scan line and slope algorithms; it proposes an illegal building identification technique based on improved DBSCAN point cloud clustering segmentation and feature construction, combining DBSCAN clustering with a grid-based partitioning method to overcome the difficulty in adjusting parameters while optimizing clustering results and achieving rapid segmentation; and it proposes a feature-based method for constructing an illegal building knowledge base. Verification of the illegal building screening effect on the verification road section shows that this invention has high practical value and good effectiveness, meets the needs of actual engineering applications, and plays an important auxiliary role in the rapid and automated screening of illegal buildings on ordinary trunk roads.

[0076] Example 2:

[0077] The computer-readable storage medium of this embodiment stores a computer program that, when executed by a processor, implements the steps in the automated investigation method for illegal constructions on ordinary trunk highways based on laser point clouds in Embodiment 1.

[0078] The computer-readable storage medium in this embodiment can be an internal storage unit of the terminal, such as the terminal's hard disk or memory; the computer-readable storage medium in this embodiment can also be an external storage device of the terminal, such as a plug-in hard disk, smart memory card, secure digital card, flash memory card, etc. equipped on the terminal; furthermore, the computer-readable storage medium can include both the terminal's internal storage unit and external storage devices.

[0079] The computer-readable storage medium of this embodiment is used to store computer programs and other programs and data required by the terminal. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0080] Example 3:

[0081] The computer device of this embodiment includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the automated investigation method for illegal construction on ordinary trunk highways based on laser point clouds in Embodiment 1.

[0082] In this embodiment, the processor can be a central processing unit, or other general-purpose processors, digital signal processors, application-specific integrated circuits, off-the-shelf programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc. The memory can include read-only memory and random access memory, and provides instructions and data to the processor. A portion of the memory can also include non-volatile random access memory. For example, the memory can also store device type information.

[0083] Those skilled in the art will understand that the content disclosed in the embodiments can be provided as a method, system, or computer program product. Therefore, this solution can take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this solution can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage) containing computer-usable program code.

[0084] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. An automated method for detecting illegal constructions on ordinary trunk highways based on laser point clouds, characterized in that, Includes the following steps: Step A: Based on the vehicle-mounted laser point cloud scanning data, extract the geometric boundaries of ordinary trunk highways and select the highway land area. Step B: Calculate the relative distance between the point cloud of buildings / structures and the edge line, and extract the laser point cloud within the highway and highway building control area; Remove gross errors and local outliers from the original point cloud data to achieve overall effective point cloud extraction of illegal construction objects; Step C: Based on the improved DBSCAN point cloud clustering and segmentation, the filtered non-ground point cloud is clustered and segmented to divide each ground feature point cloud into independent illegal building objects to be identified. Step D: Construct a knowledge base of illegal constructions based on features and identify the types of illegal constructions; Step E: Based on the ATD file and the point cloud recognition results of the illegal buildings, perform image re-recognition of the illegal buildings; In step C, the point cloud clustering and segmentation based on the improved DBSCAN includes the following steps: Step C1, based on mesh generation: (1) Based on the distribution characteristics of point clouds and data processing requirements, establish a one-to-one correspondence between the serial number and coordinate value of spatial points, and set the filtered point cloud; (2) After the index is established, the three-dimensional range is confirmed and divided into equal intervals. The three-dimensional grid is formed by dividing the grid with a 1-meter interval in the direction of vehicle travel, a 0.2-meter interval in the vertical direction, and a 0.2-meter interval in the height direction. (3) Based on the three-dimensional information of the points, match them into their respective grids to form corresponding numbers; Step C2, DBSCAN clustering and segmentation: (1) Based on the gridding, labels are added to each point in the point cloud according to the point density in the grid to achieve segmentation; (2) Extract the points of each label from the overall point cloud based on the labels. The point cloud composed of the points of each label will correspond to an object to be identified, which will serve as the basis for subsequent point cloud recognition. Step D, which involves constructing a knowledge base for illegal constructions and identifying types of illegal structures based on features, includes the following steps: Step D1: For illegal roadside structures, the following procedure shall be followed for identification: (1) Retrieve the road edge point A that is closest to the point O where the point cloud of the building / structure contacts the ground, search for the road edge point on the opposite side, retrieve the road edge point B that is closest to point A, and determine whether the building is inside or outside the road based on the coordinates of the ground points A, B, and O. (2) If it is located within the edge of the highway, it is considered an illegal structure; (3) If it is outside the edge line of the highway, calculate and determine the minimum distance between OA and OB; (4) If the minimum distance is less than the distance requirement of the highway construction control zone, it shall be determined as an illegal building; otherwise, it shall be determined as a non-illegal building. Step D2: For illegal overhead pipelines, the following procedure shall be followed for identification: (1) Retrieve the road surface point A that is closest to the lowest point O of the overhead object point cloud, calculate the distance between OA, and determine whether the overhead object is within or outside the highway construction clearance: (2) If the distance between OA is less than the road clearance distance requirement, it shall be judged as an illegal building; otherwise, it shall be judged as a non-illegal building. (3) For overhead transmission lines, determine whether the distance between OA is less than the minimum vertical distance requirement from the road surface at the intersection of the transmission line conductor and the highway. If it is less, it is determined to be an illegal construction; otherwise, it is determined to be a non-illegal construction. Step D3: Based on the spatial topological relationship, physical size, and point cloud characteristics of different illegal constructions and the ground, an illegal construction knowledge base is established to further determine the types of illegal constructions. The point cloud characteristics of three typical types of illegal constructions in the knowledge base are shown in Table 1. Table 1. Point cloud features of three typical types of illegal constructions 。 2. The automated method for detecting illegal constructions on ordinary trunk highways based on laser point clouds according to claim 1, characterized in that, In step A, point cloud elevation, point cloud density, slope difference analysis, and road edge point clustering methods are used to extract the geometric boundaries of ordinary trunk highways.

3. The automated method for detecting illegal constructions on ordinary trunk highways based on laser point clouds according to claim 2, characterized in that, In step A, the scope of highway land use is determined according to the highway construction control zone of "no less than 20 meters for national highways and no less than 15 meters for provincial highways".

4. The automated method for detecting illegal constructions on ordinary trunk highways based on laser point clouds according to claim 3, characterized in that, In step B, gross errors and local outliers in the original point cloud data are removed by combining a pass filter with a radius filter.

5. The automated method for detecting illegal constructions on ordinary trunk highways based on laser point clouds according to claim 4, characterized in that, In step E, based on the ATD file and the point cloud recognition results of illegal buildings, real-scene images of the corresponding locations of illegal buildings are extracted, and the illegal building type is identified and classified in the real-scene images through the target recognition model, thus completing the re-recognition of illegal building images.

Images