Vehicle contour detection method and system based on laser radar

Abstract

The invention discloses a vehicle contour detection method and system based on a laser radar, a high-frequency (60Hz) laser radar sensor is adopted to scan a vehicle, the system can generate high-resolution point cloud data, fine structural features of the vehicle can be captured through high-frequency scanning, the contour measurement precision is ensured, and the detection precision is improved. The point cloud data processing unit utilizes advanced algorithms (such as a kd-tree-based neighborhood averaging method and a least square method) to carry out denoising and boundary point identification, the accuracy of a measurement result is further improved, through the methods, the system can accurately identify and measure the length, the width and the height of the vehicle, errors are controlled within a minimum range, and the accuracy of the measurement result is improved. The system adopts an efficient kd-tree structure to manage and process point cloud data, the data processing speed is remarkably improved, and the kd-tree has excellent performance in the aspects of spatial query and nearest neighbor search, so that data processing is more efficient.

Description

Technical Field

[0001] The present invention relates to the field of vehicle detection technology, and in particular to a vehicle outline detection method and system based on laser radar. Background Art

[0002] With the development of intelligent transportation systems, accurate measurement of vehicle profile dimensions has become increasingly important. Traditional vehicle profile detection methods mainly rely on manual measurement or simple electronic equipment, which are not only time-consuming and inefficient, but also greatly affected by human factors and have limited accuracy. In addition, traditional methods are difficult to achieve real-time measurement and dynamic monitoring, and cannot meet the measurement requirements of complex environments and complex objects. Traditional vehicle profile detection technology mainly relies on manual measurement or systems based on simple sensors.

[0003] These methods have the following defects and shortcomings:

[0004] 1. Limited accuracy: Manual measurement is easily affected by the operator's skills, experience and subjective judgment, resulting in deviations in the measurement results;

[0005] 2. Inefficiency: Traditional methods usually require vehicles to stop at specific locations for measurement, which not only increases detection time but also affects the efficiency of traffic flow;

[0006] 3. Low degree of automation: measurement methods that rely on manual work are difficult to automate and cannot meet the needs of large-scale or continuous monitoring;

[0007] 4. Complex data processing: Simple sensor systems often require complex post-data processing to ensure the accuracy of measurement results, which increases the overall complexity of the system;

[0008] 5. Poor environmental adaptability: Under adverse weather or lighting conditions, the accuracy and reliability of traditional detection methods will drop significantly.

[0009] The development of LiDAR technology provides a new solution for measuring vehicle dimensions. LiDAR can quickly and accurately scan vehicles and generate high-precision point cloud data, thereby realizing automatic detection of vehicle dimensions. Therefore, how to effectively use LiDAR technology to improve measurement accuracy and efficiency, and to perform dynamic measurement when the vehicle passes through the detection area without stopping, has become a problem that needs to be solved urgently.

[0010] In view of this, a vehicle outline detection method and system based on laser radar are designed. Summary of the invention

[0011] The purpose of the present invention is to provide a vehicle outline detection method and system based on laser radar to solve the problems existing in the prior art mentioned in the above background technology.

[0012] To achieve the above object, the present invention provides the following technical solution: a vehicle outline detection system based on laser radar, comprising:

[0013] At least one laser radar sensor with a scanning frequency of 60 Hz, used to scan vehicles passing through the detection area and generate point cloud data of the vehicle's outline;

[0014] a data processing unit, including a point cloud data denoising module and a boundary point identification module, for receiving and processing point cloud data to determine the outer dimensions of the vehicle;

[0015] And a user interface for displaying the measurement results of vehicle dimensions and generating PDF reports.

[0016] Preferably, the point cloud data denoising module uses a kd-tree-based neighborhood averaging method to perform data denoising, and the boundary point identification module uses a least squares method to determine local curvature features to protect boundary features.

[0017] Preferably, the installation height of the laser radar is 4800-5300 mm, and the scanning frequency of the laser radar is 60 Hz.

[0018] Preferably, a plurality of sensors are further included to assist the laser radar sensor in accurately measuring the vehicle profile, and the specific steps include:

[0019] A1) Use auxiliary sensors to determine whether a vehicle has entered the detection area;

[0020] A2) After the vehicle enters the detection area, the lidar sensor is activated for scanning;

[0021] A3) Collect the point cloud data generated by the LiDAR sensor and fuse it with the auxiliary sensor data;

[0022] A4) Improve the accuracy and reliability of vehicle external dimension measurement through fused data.

[0023] Preferably, the boundary point identification module of the data processing unit specifically includes the following steps:

[0024] B1) Collect point cloud data generated by the LiDAR sensor;

[0025] B2) Establish the spatial topological relationship of point cloud data based on kd-tree;

[0026] B3) Use neighborhood averaging method to denoise point cloud data;

[0027] B4) Using the least squares method to determine the local curvature features in the point cloud data;

[0028] B5) Protect the boundary features of the vehicle outline through a boundary point recognition algorithm.

[0029] A vehicle profile detection system based on laser radar, the detection method of which comprises the following steps:

[0030] S1) determining whether the vehicle enters the detection area;

[0031] S2) scanning the vehicle using a lidar sensor and generating point cloud data;

[0032] S3) processing the point cloud data to determine the outer dimensions of the vehicle;

[0033] S4) outputs the measurement results of the vehicle's external dimensions.

[0034] Preferably, step S1 specifically includes:

[0035] S11) acquiring a radar point cloud data frame corresponding to each moment from the laser radar sensor;

[0036] S12) setting a Y-axis direction threshold and filtering coordinate points to remove points that satisfy Yi<h0;

[0037] S13) merging the left and right radar point cloud data frames to find the minimum and maximum values ​​of the X coordinates;

[0038] S14) When the width of the object in the detection area is greater than a threshold, it is determined that the vehicle has entered the detection area.

[0039] Preferably, step S2 specifically includes:

[0040] S21) collecting a frame of point cloud data transmitted from the front laser radar when the vehicle leaves the detection area;

[0041] S22) converting the coordinate value of each point in the frame point cloud data;

[0042] S23) searching for the point with the smallest X-axis coordinate value in the frame point cloud data;

[0043] S24) Set the horizontal distance between the front laser radar sensor and the left and right laser radar scanning planes to L, and the calculation formula for the vehicle length is: L Car =Ls.

[0044] Preferably, step S3 specifically includes:

[0045] S31) Calculate the maximum X coordinate value (x{max}) and the minimum X coordinate value (x{min}) of all point cloud data, and the calculation formula of vehicle width (W) is: W = x max -x min ;

[0046] S32) Calculate the maximum Y coordinate value (y{max}) and the minimum Y coordinate value (y{min}) of all point cloud data, and the calculation formula of vehicle height (H) is: H=y max -y min .

[0047] Preferably, step S4 further includes using a user interface to display the measurement results and generate a PDF report of the vehicle's external dimensions.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. High-precision contour measurement: By using a high-frequency (60Hz) lidar sensor to scan the vehicle, the system can generate high-resolution point cloud data. High-frequency scanning can capture the subtle structural features of the vehicle, ensuring the accuracy of contour measurement. The point cloud data processing unit uses advanced algorithms (such as the neighborhood average method and least squares method based on kd-tree) for denoising and boundary point identification, further improving the accuracy of the measurement results. Through these methods, the system can accurately identify and measure the length, width and height of the vehicle, and the error is controlled within a very small range.

[0050] 2. Efficient data processing and real-time feedback: The system uses an efficient kd-tree structure to manage and process point cloud data, which significantly improves the data processing speed. Kd-tree has superior performance in spatial query and nearest neighbor search, making data processing more efficient. Real-time point cloud data denoising and boundary point recognition algorithms can process data in real time as the vehicle passes through the detection area and quickly output measurement results. The user interface can display the vehicle's outer dimensions in real time, providing users with instant feedback.

[0051] 3. Multi-sensor collaboration: The system supports multiple LiDAR sensors to work together to scan the vehicle from different angles and generate more comprehensive point cloud data. This multi-sensor collaborative working mode can reduce blind spots and improve the comprehensiveness and reliability of detection. The collaborative work of multiple sensors can also capture more detailed information when the vehicle passes through the detection area, further improving the accuracy of contour detection.

[0052] 4. Intelligent denoising and boundary protection: By denoising point cloud data based on the kd-tree neighborhood average method, the system can effectively remove noise points and retain the main structural features of the data. This process can significantly improve the quality of data and reduce errors. The boundary point recognition module uses the least squares method to determine the local curvature characteristics and protects the boundary features of the vehicle outline through the boundary point recognition algorithm. This intelligent boundary protection mechanism ensures the accuracy and integrity of the boundary features and avoids accidental deletion or modification during data processing.

[0053] 5. User-friendly and multi-functional interface: The system has a user-friendly design, which can display the vehicle's external dimensions in real time and generate detailed PDF reports. These reports record each measurement step and data, providing users with a comprehensive and intuitive reference.

[0054] 6. Wide range of application scenarios: This system can be widely used in logistics companies, traffic management departments and other scenarios. For logistics companies, the system can help them manage their fleets efficiently and ensure that the vehicle size meets the transportation standards; for traffic management departments, the system can be used to detect large vehicles on the road to prevent overload problems. The high precision and high efficiency of the system enable it to maintain reliable performance in various complex detection environments, providing strong technical support for different application scenarios.

[0055] 7. Strong integration and scalability: The system design has good integration and scalability, and can be easily integrated with the existing traffic management system or logistics management system. Users can flexibly configure and expand system functions according to their needs to meet different application requirements. Remote access and management functions make the system easy to maintain and upgrade. Users can remotely monitor and operate the system through the network at any time, which improves the convenience and efficiency of management.

[0056] To sum up, the laser radar-based vehicle profile detection system described in this patent has significant advantages in high precision, high efficiency, multi-function and wide application, and provides an advanced and reliable technical solution for vehicle profile detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A flowchart of the steps of assisting a laser radar sensor in accurately measuring the outer contour of a vehicle according to the present invention;

[0058] Figure 2 A flow chart of the steps of the boundary point identification module of the present invention;

[0059] Figure 3 The present invention is a flowchart of the steps of a vehicle outline detection method based on laser radar. DETAILED DESCRIPTION

[0060] 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.

[0061] In the description of the invention, it should be noted that the terms "upper", "lower", "inner", "outer", "front end", "rear end", "two ends", "one end", "the other end" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.

[0062] In the description of the invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "provided with", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0063] See also Figure 1-3 The present invention provides a technical solution: a vehicle outline detection system based on laser radar, characterized in that it includes:

[0064] At least one laser radar sensor with a scanning frequency of 60 Hz, used to scan vehicles passing through the detection area and generate point cloud data of the vehicle's outline;

[0065] a data processing unit, including a point cloud data denoising module and a boundary point identification module, for receiving and processing point cloud data to determine the outer dimensions of the vehicle;

[0066] And a user interface for displaying the measurement results of vehicle dimensions and generating PDF reports.

[0067] Specifically, the vehicle contour detection system based on LiDAR can accurately capture and generate the vehicle contour point cloud data by using a high-frequency scanning LiDAR sensor. This high-frequency scanning (60Hz) can ensure that fast-moving vehicles will not cause data loss when passing through the detection area, thereby ensuring the integrity and accuracy of the data. Through the point cloud data denoising module and boundary point recognition module of the data processing unit, noise points can be effectively eliminated, the boundary features of the vehicle contour can be protected, and the reliability of the detection results can be ensured. The user interface provides a friendly operating platform for users to view and analyze the vehicle contour dimensions in real time, and can generate PDF reports for subsequent archiving and analysis. The overall system design is simple and efficient, can adapt to a variety of complex detection environments, and has high practical value and market potential.

[0068] Among them, the point cloud data denoising module uses the neighborhood averaging method based on kd-tree for data denoising, and the boundary point recognition module uses the least squares method to judge the local curvature features to protect the boundary features.

[0069] Specifically, the point cloud data denoising module uses the neighborhood averaging method based on kd-tree for data denoising, which can effectively reduce the noise in the point cloud data and retain useful structural information. This method uses the spatial topological relationship of kd-tree and smoothes the point cloud data through the neighborhood averaging method. It can remove most of the noise points while retaining the boundary features, significantly improving the clarity and accuracy of the data. The boundary point recognition module uses the least squares method to determine the local curvature features, which can accurately locate the boundary points of the vehicle outline. Through these steps, the system can more accurately capture and identify the true size of the vehicle outline and provide more reliable detection results. This design improves the robustness and accuracy of the system, making it more effective in practical applications.

[0070] The installation height of the laser radar is 4800-5300mm, and the scanning frequency of the laser radar is 60Hz.

[0071] Specifically, the scanning angle range of the laser radar is 180 degrees, and can be adjusted to adapt to different detection environments and requirements.

[0072] Specifically, the installation height of the LiDAR is within the range of 4800-5300mm, and combined with its 180-degree scanning angle, it can cover a larger detection area. This height and angle configuration can effectively reduce blind spots and improve the detection coverage and accuracy of vehicles of different sizes. Adjusting the scanning angle to adapt to different detection environments and requirements gives the system greater flexibility and adaptability. Whether it is a tall vehicle or a short vehicle, the system can effectively detect and adapt to the needs of different scenarios. With this configuration, the system significantly improves adaptability and flexibility while ensuring high accuracy and high reliability.

[0073] Further comprising a plurality of sensors for assisting the laser radar sensor in accurately measuring the vehicle profile, the specific steps comprising:

[0074] A1) Use auxiliary sensors to determine whether a vehicle has entered the detection area;

[0075] A2) After the vehicle enters the detection area, the lidar sensor is activated for scanning;

[0076] A3) Collect the point cloud data generated by the LiDAR sensor and fuse it with the auxiliary sensor data;

[0077] A4) Improve the accuracy and reliability of vehicle external dimension measurement through fused data.

[0078] Specifically, the auxiliary configuration of multiple sensors can significantly improve the ability to accurately measure the vehicle's profile. Scanning at multiple angles can reduce the blind spots and data loss problems that may exist at a single angle, thereby improving the comprehensiveness and accuracy of detection. By simultaneously scanning the vehicle from different angles with multiple sensors, more complete and accurate point cloud data can be obtained to ensure the accuracy and reliability of the detection results. This multi-sensor configuration not only improves the detection accuracy of the system, but also enhances the robustness and stability of the system, and can adapt to more complex detection environments and application scenarios.

[0079] The boundary point recognition module of the data processing unit specifically includes the following steps:

[0080] B1) Collect point cloud data generated by the LiDAR sensor;

[0081] B2) Establish the spatial topological relationship of point cloud data based on kd-tree;

[0082] B3) Use neighborhood averaging method to denoise point cloud data;

[0083] B4) Using the least squares method to determine the local curvature features in the point cloud data;

[0084] B5) Protect the boundary features of the vehicle outline through a boundary point recognition algorithm.

[0085] Specifically, in step B1, by collecting point cloud data generated by high-frequency scanning, the system can obtain detailed and accurate vehicle contour information. This step ensures the integrity and high resolution of the data. In step B2, the kd-tree structure is used to organize and process point cloud data, which can effectively manage and query the spatial relationship between data points. The efficiency and superior query performance of kd-tree can significantly improve the speed and accuracy of point cloud data processing. In step B3, the system can smooth the point cloud data and remove isolated noise points and irregular data through the neighborhood averaging method. This method can retain the main structural features of the point cloud data while eliminating noise interference, significantly improving the quality and accuracy of the data. After denoising, the data is clearer and cleaner, providing a more reliable basis for subsequent boundary point identification. In step B4, the least squares method can accurately calculate the local curvature features of the point cloud data and identify the boundary points of the vehicle contour. This method can find a balance between high precision and low computational cost, ensuring the accuracy and efficiency of boundary identification. By judging the curvature features, the system can accurately locate and describe the vehicle's contour boundary. In step B5, after identifying the boundary points, the boundary point recognition algorithm further protects these boundary features to prevent them from being accidentally deleted or modified in subsequent processing steps. This protection mechanism can ensure the integrity and accuracy of the final detection results and provide more reliable vehicle outline dimension measurement results.

[0086] Specifically, through the above steps B1-B5, the system can efficiently and accurately process and analyze the point cloud data generated by the lidar, and accurately identify the outer contour boundary features of the vehicle. This process not only improves the efficiency and accuracy of data processing, but also enhances the robustness and stability of the system. It can maintain reliable performance in various complex detection environments, and significantly improves the overall effect of vehicle outline detection.

[0087] A vehicle profile detection system based on laser radar, the detection method of which comprises the following steps:

[0088] S1) determining whether the vehicle enters the detection area;

[0089] S2) scanning the vehicle using a lidar sensor and generating point cloud data;

[0090] S3) processing the point cloud data to determine the outer dimensions of the vehicle;

[0091] S4) outputs the measurement results of the vehicle's external dimensions.

[0092] Specifically, the vehicle outline detection method using LiDAR can complete the vehicle outline detection in a systematic and standardized manner through multiple steps. First, determine whether the vehicle enters the detection area to ensure the orderly progress of the detection process. Then, use the LiDAR sensor to perform high-frequency scanning to generate accurate point cloud data. Next, process the point cloud data to determine the vehicle's outline dimensions to ensure the effectiveness and accuracy of data processing. Finally, output the measurement results of the vehicle's outline dimensions to provide intuitive detection feedback. This method has clear steps and a reasonable process, which can effectively improve the efficiency and accuracy of detection and provide reliable technical support for practical applications.

[0093] Among them, step S1 specifically includes:

[0094] S11) acquiring a radar point cloud data frame corresponding to each moment from the laser radar sensor;

[0095] S12) setting a Y-axis direction threshold and filtering coordinate points to remove points that satisfy Yi<h0;

[0096] S13) merging the left and right radar point cloud data frames to find the minimum and maximum values ​​of the X coordinates;

[0097] S14) When the width of the object in the detection area is greater than a threshold, it is determined that the vehicle has entered the detection area.

[0098] Specifically, in the step of determining whether a vehicle has entered the detection area, by obtaining the point cloud data frame of the lidar sensor, setting the Y-axis threshold and filtering the coordinate points, false detection and missed detection can be effectively reduced. By merging the left and right radar point cloud data frames and finding the minimum and maximum values ​​of the X coordinates, it is possible to more accurately determine whether the vehicle has entered the detection area. In this way, the system can more accurately determine the presence of the vehicle, reduce false positives, and improve the accuracy and efficiency of detection. This method not only improves the detection accuracy of the system, but also simplifies the operation process and facilitates practical application.

[0099] Among them, step S2 specifically includes:

[0100] S21) collecting a frame of point cloud data transmitted from the front laser radar when the vehicle leaves the detection area;

[0101] S22) converting the coordinate value of each point in the frame point cloud data;

[0102] S23) searching for the point with the smallest X-axis coordinate value in the frame point cloud data;

[0103] S24) Set the horizontal distance between the front laser radar sensor and the left and right laser radar scanning planes to L, and the calculation formula for the vehicle length is: L Car =Ls.

[0104] Specifically, in the step of scanning the vehicle and generating point cloud data, by collecting a frame of point cloud data transmitted by the front laser radar when the vehicle leaves the detection area and converting the coordinate value of each point in the frame of point cloud data, the size information of the vehicle can be captured more accurately. In the frame of point cloud data, find the point with the smallest X-axis coordinate value, set the horizontal distance between the front laser radar sensor and the left and right laser radar scanning planes to L, and use the calculation formula of the vehicle length to accurately calculate the length of the vehicle. This method ensures the accuracy and reliability of the detection results through specific data processing and calculation steps.

[0105] The S3 step specifically includes:

[0106] S31) Calculate the maximum X coordinate value (x{max}) and the minimum X coordinate value (x{min}) of all point cloud data, and the calculation formula of vehicle width (W) is: W = x max -x min ;

[0107] S32) Calculate the maximum Y coordinate value (y{max}) and the minimum Y coordinate value (y{min}) of all point cloud data, and the calculation formula of vehicle height (H) is: H=y max -y min .

[0108] Specifically, in the step of processing point cloud data to determine the outer dimensions of the vehicle, the width of the vehicle can be accurately calculated by calculating the maximum X coordinate value and the minimum X coordinate value in all point cloud data; the height of the vehicle can be accurately calculated by calculating the maximum Y coordinate value and the minimum Y coordinate value in all point cloud data. By preprocessing the point cloud data using artificial intelligence algorithms, noise points can be automatically identified and removed, improving the efficiency and accuracy of boundary point identification. This method significantly improves the accuracy and efficiency of detection through detailed data processing and calculation steps combined with artificial intelligence algorithms.

[0109] The S4 step further includes displaying the measurement results using a user interface and generating a PDF report of the vehicle's external dimensions.

[0110] Specifically, the system further includes remote access and management functions, allowing users to remotely view and download the measurement results of the vehicle's external dimensions through the network; the user interface includes a real-time display module for displaying the real-time changes in the vehicle's external dimensions, a historical data query module for querying and comparing past measurement results, and supports a report generation module to customize reports in different formats according to user needs and support multiple languages.

[0111] Specifically, in the step of outputting the measurement results of the vehicle's external dimensions, the measurement results are displayed through the user interface, and a PDF report of the vehicle's external dimensions is generated, which can intuitively display the test results to the user and provide a detailed test report. The measurement results of the vehicle's external dimensions are sent to the cloud server through the network interface, which facilitates centralized management and remote access, and can improve the convenience and security of data management. This step not only provides a friendly user interface and detailed test reports, but also improves the efficiency and security of data management, and provides users with comprehensive testing services.

[0112] Specifically, through the remote access and management functions, users can remotely view and download the measurement results of the vehicle's external dimensions through the network, which improves the convenience and practicality of the system. The user interface includes a real-time display module for displaying the real-time changes in the vehicle's external dimensions, a historical data query module for querying and comparing past measurement results, and a report generation module that customizes reports in different formats according to user needs and supports multiple languages. This design not only improves the user experience and operational convenience of the system, but also enhances the adaptability and functionality of the system, providing users with a comprehensive and flexible detection solution.

[0113] Example 1

[0114] Implementation background:

[0115] A logistics company needs to accurately measure the outer dimensions of its fleet vehicles in order to optimize its logistics operations and garage management. The company decided to use the above-mentioned LiDAR-based vehicle outer dimensions detection system to detect the outer dimensions of its fleet vehicles.

[0116] Implementation steps:

[0117] 1. System setup and initialization:

[0118] A lidar sensor is installed at the entrance of the logistics company's garage, with an installation height of 5000mm and a scanning angle set to 180 degrees.

[0119] The scanning frequency of the lidar sensor is configured to be 60Hz, ensuring that the contour data of the vehicle can be captured quickly and accurately when it passes through the detection area.

[0120] A point cloud data denoising module and a boundary point recognition module are configured in the data processing unit to ensure that the point cloud data can be effectively processed.

[0121] 2. Data collection:

[0122] When a truck measuring 8 meters long, 2.5 meters wide and 3.2 meters high enters the detection area, the lidar sensor starts working.

[0123] The system uses a lidar sensor to scan the vehicle at a frequency of 60Hz to generate high-precision point cloud data.

[0124] 3. Data processing:

[0125] Point cloud data denoising: Use the kd-tree-based neighborhood averaging method to denoise the point cloud data, remove noise points, and retain useful structural information.

[0126] Establish spatial topological relationships: Establish spatial topological relationships of point cloud data based on kd-tree to provide a basis for subsequent boundary point identification.

[0127] Boundary point recognition: Use the least squares method to determine the local curvature features in the point cloud data and identify the boundary points of the vehicle outline.

[0128] 4. Dimension calculation:

[0129] By calculating the maximum X coordinate value and the minimum X coordinate value in the point cloud data, it is determined that the width of the vehicle is 2.5 meters.

[0130] By calculating the maximum Y coordinate value and the minimum Y coordinate value in the point cloud data, it is determined that the height of the vehicle is 3.2 meters.

[0131] Based on the point cloud data when the vehicle enters and leaves the detection area, the calculation formula L Car =Ls, the calculated length of the vehicle is 8 meters.

[0132] 5. Result output:

[0133] The system displays the measurement results of the vehicle's external dimensions in real time through the user interface.

[0134] Generates a detailed PDF report of the vehicle's dimensions, including the length, width and height of the vehicle, as well as detailed data for each measurement step.

[0135] The measurement results are sent to the cloud server via the network interface for centralized management and remote access.

[0136] 6. Data management and query:

[0137] Users can access the system remotely to view and download the measurement results of vehicle dimensions in real time.

[0138] The system supports historical data query, and users can compare measurement results in different time periods and analyze changes in vehicle dimensions.

[0139] Supports report generation in multiple languages ​​and formats, and customizes report content according to user needs.

[0140] Implementation results:

[0141] Through the above implementation steps, the logistics company successfully measured the external dimensions of its fleet vehicles. The measurement results showed that the length of the vehicle was 8 meters, the width was 2.5 meters, and the height was 3.2 meters, with an error within 0.5%. The generated PDF report recorded each measurement step and data in detail, providing an important basis for subsequent vehicle management. Through the remote access function, users can view and manage measurement data anytime and anywhere, improving the convenience and efficiency of management.

[0142] Example 2

[0143] Implementation background:

[0144] The traffic management department of a certain city needs to detect the outer profile of large trucks on the city roads to ensure that the vehicles comply with road traffic regulations and avoid overloading. The department decided to use the above-mentioned LiDAR-based vehicle outer profile detection system to detect trucks passing through a certain main road.

[0145] Implementation steps:

[0146] 1. System setup and initialization:

[0147] Multiple lidar sensors are installed at checkpoints on major roads, with mounting heights of 4900mm and 5200mm respectively, and the scanning angle is set to 180 degrees.

[0148] Multiple lidar sensors are configured to work simultaneously to scan the vehicle from different angles to ensure the comprehensiveness and accuracy of the data.

[0149] A point cloud data denoising module and a boundary point recognition module are configured in the data processing unit to ensure that the point cloud data can be effectively processed.

[0150] 2. Data collection:

[0151] When a truck measuring 10 meters long, 2.8 meters wide and 3.5 meters high enters the detection area, the lidar sensor starts working.

[0152] The system uses multiple lidar sensors to scan the vehicle at a frequency of 60Hz to generate high-precision point cloud data.

[0153] 3. Data processing:

[0154] Point cloud data denoising: Use the kd-tree-based neighborhood averaging method to denoise the point cloud data, remove noise points, and retain useful structural information.

[0155] Establish spatial topological relationships: Establish spatial topological relationships of point cloud data based on kd-tree to provide a basis for subsequent boundary point identification.

[0156] Boundary point recognition: Use the least squares method to determine the local curvature features in the point cloud data and identify the boundary points of the vehicle outline.

[0157] 4. Dimension calculation:

[0158] By calculating the maximum X coordinate value and the minimum X coordinate value in the point cloud data, it is determined that the width of the vehicle is 2.8 meters.

[0159] By calculating the maximum Y coordinate value and the minimum Y coordinate value in the point cloud data, it is determined that the height of the vehicle is 3.5 meters.

[0160] Based on the point cloud data when the vehicle enters and leaves the detection area, the calculation formula L Car =Ls, the calculated length of the vehicle is 10 meters.

[0161] 5. Result output:

[0162] The system displays the measurement results of the vehicle's external dimensions in real time through the user interface.

[0163] Generates a detailed PDF report of the vehicle's dimensions, including the length, width and height of the vehicle, as well as detailed data for each measurement step.

[0164] The measurement results are sent to the traffic management department’s cloud server via a network interface for centralized management and remote access.

[0165] 6. Data management and query:

[0166] Users can access the system remotely to view and download the measurement results of vehicle dimensions in real time.

[0167] The system supports historical data query, and users can compare measurement results in different time periods and analyze changes in vehicle dimensions.

[0168] Supports report generation in multiple languages ​​and formats, and customizes report content according to user needs.

[0169] Implementation results:

[0170] Through the above implementation steps, the traffic management department successfully carried out accurate measurement of trucks passing through the main roads. The measurement results showed that the length of the vehicle was 10 meters, the width was 2.8 meters, and the height was 3.5 meters, with an error within 0.3%. The generated PDF report recorded each measurement step and data in detail, providing an important basis for traffic management. Through the remote access function, users can view and manage measurement data anytime and anywhere, improving the convenience and efficiency of management. The system effectively helps the traffic management department monitor the outer dimensions of large trucks on the road and avoid the occurrence of over-limit and overload problems.

[0171] Example 3

[0172] In a large logistics center, a LiDAR-based vehicle profile detection system is installed at the entrance and exit inspection stations. The purpose of the system is to automatically measure the profile dimensions of trucks entering and leaving the logistics center to ensure that they meet the safety standards for road transportation.

[0173] System Configuration:

[0174] LiDAR sensor: Two LiDAR sensors with a scanning frequency of 60 Hz are installed at a height of 5000 mm and a coverage detection area width of 10 m.

[0175] Data processing unit: Equipped with a high-performance computer system and installed with specialized software for receiving and processing point cloud data.

[0176] User Interface: A user-friendly interface is designed to display measurement results and generate PDF reports.

[0177] Steps:

[0178] When a truck approaches the inspection station, the pressure sensor on the ground is triggered and the system is automatically activated.

[0179] The LiDAR sensor starts scanning, continuously generating point cloud data.

[0180] The data processing unit receives the point cloud data and first uses the kd-tree algorithm to denoise the data.

[0181] Next, the least squares method is used to identify boundary points of the denoised data to determine the outline of the vehicle.

[0182] The user interface displays the 3D vehicle model and dimensional data in real time, and the operator can monitor the measurement process.

[0183] Measurement results:

[0184] Vehicle length: The measured length of the truck is 12.5m, with an error range of ±5cm.

[0185] Vehicle width: The measured width of the truck is 2.55m, with an error range of ±5cm.

[0186] Vehicle height: The measured height of the truck is 3.8m, with an error range of ±5cm.

[0187] Data processing and display:

[0188] The measurement results are automatically recorded in the system database and a detailed PDF report is generated.

[0189] The report includes a 3D model image of the vehicle, dimensional data, and measurement time and date.

[0190] The system also provides remote access capabilities, allowing managers to remotely view and download measurement results via the network.

[0191] Embodiment 4:

[0192] In an automobile manufacturing plant, a lidar-based vehicle profile detection system was introduced to ensure that newly produced cars met dimensional specifications.

[0193] System Configuration:

[0194] LiDAR sensor: Four LiDAR sensors with a scanning frequency of 60 Hz are installed at a height of 5200 mm and a coverage detection area width of 15 m.

[0195] Data processing unit: Equipped with advanced data processing server and installed with the latest point cloud processing software.

[0196] User Interface: An interactive touch screen interface was developed to display the measurement results and generate PDF reports.

[0197] Steps:

[0198] When a newly manufactured car enters the detection area, the infrared sensor is triggered and the system starts automatically.

[0199] The LiDAR sensor starts scanning and continuously generates point cloud data at 1 second intervals.

[0200] The data processing unit receives the point cloud data, performs denoising using the kd-tree algorithm, and uses the least squares method to identify boundary points.

[0201] The user interface displays the 3D model and dimensional data of the car in real time, and the operator can monitor the measurement process in real time.

[0202] Measurement results:

[0203] Car length: The measured car length is 4.7m, with an error range of ±2cm.

[0204] Car width: The measured car width is 1.8m, with an error range of ±2cm.

[0205] Vehicle height: The measured vehicle height is 1.4m, with an error range of ±2cm.

[0206] Data processing and display:

[0207] The measurement results are automatically recorded in the system database and a detailed PDF report is generated.

[0208] The report includes an image of the 3D model of the car, dimensional data, and the time and date of measurement.

[0209] The system also provides remote access capabilities, allowing quality control personnel to remotely view and download measurement results via the network.

[0210] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A vehicle outline detection system based on laser radar, characterized in that: include: At least one laser radar sensor with a scanning frequency of 60 Hz, used to scan vehicles passing through the detection area and generate point cloud data of the vehicle's outline; a data processing unit, including a point cloud data denoising module and a boundary point identification module, for receiving and processing point cloud data to determine the outer dimensions of the vehicle; And a user interface for displaying the measurement results of vehicle dimensions and generating PDF reports.

2. The vehicle outline detection system based on laser radar according to claim 1, characterized in that: in, The point cloud data denoising module uses the neighborhood averaging method based on kd-tree for data denoising, and the boundary point recognition module uses the least squares method to determine the local curvature features to protect the boundary features.

3. The vehicle outline detection system based on laser radar according to claim 1, characterized in that: The installation height of the laser radar is 4800-5300mm, and the scanning frequency of the laser radar is 60Hz.

4. The vehicle outline detection system based on laser radar according to claim 1, characterized in that: Further comprising a plurality of sensors for assisting the laser radar sensor in accurately measuring the vehicle profile, the specific steps comprising: A1) Use auxiliary sensors to determine whether a vehicle has entered the detection area; A2) After the vehicle enters the detection area, the lidar sensor is activated for scanning; A3) Collect the point cloud data generated by the LiDAR sensor and fuse it with the auxiliary sensor data; A4) Improve the accuracy and reliability of vehicle external dimension measurement through fused data.

5. The vehicle outline detection system based on laser radar according to claim 1, characterized in that: The boundary point recognition module of the data processing unit specifically includes the following steps: B1) Collect point cloud data generated by the LiDAR sensor; B2) Establish the spatial topological relationship of point cloud data based on kd-tree; B3) Use neighborhood averaging method to denoise point cloud data; B4) Using the least squares method to determine the local curvature features in the point cloud data; B5) Protect the boundary features of the vehicle outline through a boundary point recognition algorithm.

6. According to the laser radar-based vehicle outline detection system of claim 1, the detection method thereof comprises the following steps: S1) determining whether the vehicle enters the detection area; S2) scanning the vehicle using a lidar sensor and generating point cloud data; S3) processing the point cloud data to determine the outer dimensions of the vehicle; S4) outputs the measurement results of the vehicle's external dimensions.

7. The vehicle outline detection method based on laser radar according to claim 1, It is characterized in that Among them, step S1 specifically includes: S11) acquiring a radar point cloud data frame corresponding to each moment from the laser radar sensor; S12) setting a Y-axis direction threshold and filtering coordinate points to remove points that satisfy Yi<h0; S13) merging the left and right radar point cloud data frames to find the minimum and maximum values ​​of the X coordinates; S14) When the width of the object in the detection area is greater than a threshold, it is determined that the vehicle has entered the detection area.

8. The method and system for detecting vehicle outline based on laser radar according to claim 1, characterized in that: in, The S2 step specifically includes: S21) collecting a frame of point cloud data transmitted from the front laser radar when the vehicle leaves the detection area; S22) converting the coordinate value of each point in the frame point cloud data; S23) searching for the point with the smallest X-axis coordinate value in the frame point cloud data; S24) Set the horizontal distance between the front laser radar sensor and the left and right laser radar scanning planes to L, and the calculation formula for the vehicle length is: L Car =Ls.

9. The method and system for detecting vehicle outline based on laser radar according to claim 1, characterized in that: in, The S3 step specifically includes: S31) Calculate the maximum X coordinate value (x{max}) and the minimum X coordinate value (x{min}) of all point cloud data, and the calculation formula of vehicle width (W) is: W = x max -x min ; S32) Calculate the maximum Y coordinate value (y{max}) and the minimum Y coordinate value (y{min}) of all point cloud data, and the calculation formula of vehicle height (H) is: H=y max -y min .

10. The method and system for detecting vehicle outline based on laser radar according to claim 1, characterized in that: in, Step S4 further includes displaying the measurement results using a user interface and generating a PDF report of the vehicle's external dimensions.

Images