Metro gauge space detection method, device and equipment and storage medium

Abstract

The application provides a subway clearance space detection method, device and equipment and a storage medium. The method comprises the following steps: obtaining point cloud data of a subway driving environment as first point cloud data, and performing structured processing to obtain a structured map; obtaining initial pose data of the subway, calculating trajectory pose data, and extracting initial clearance space from the structured map; obtaining point cloud data of a real-time driving environment of the subway as second point cloud data, and extracting feature point cloud of the second point cloud data; calculating inter-frame pose data of the subway; calculating real-time pose data of the subway based on the inter-frame pose data, the trajectory pose data and the structured map; calculating an offset of the real-time pose data relative to the trajectory pose data, and calculating real-time clearance space. Through the method of pre-building a structured map and pre-extracting initial clearance space, the operation amount of detecting real-time clearance space is reduced, and the real-time performance and operation safety are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rail transit detection, and in particular to a subway clearance space detection method, device, equipment and storage medium. BACKGROUND

[0002] Clearance refers to the contour size line that is not allowed to be exceeded for the purpose of ensuring the safety of locomotive vehicles running on railway lines and preventing locomotive vehicles from colliding with adjacent line buildings and equipment. Subway clearance detection is one of important means to ensure the safe operation of subways. From the perspective of the vehicle body, the cross section of the subway is limited not to exceed the clearance, and from the perspective of the equipment and buildings, the cross section of the equipment and buildings is limited not to invade the clearance. Therefore, it has extremely important safety significance to detect the clearance space of the subway in real time and judge whether there is an invasion risk during the operation of the subway. The existing technology is mainly used for clearance space detection in the process of subway tunnel construction, and it is difficult to obtain real-time position and front clearance space in the operation of the subway. Moreover, each subway needs to be equipped with multiple sets of high-precision devices, which is difficult to be used in large-scale operation. SUMMARY

[0003] Therefore, the purpose of the present application is to provide a subway clearance space detection method, device, equipment and storage medium.

[0004] In order to achieve the above purpose, the present application provides a subway clearance space detection method, which comprises:

[0005] Obtaining point cloud data of all target objects in a subway driving environment as first point cloud data, and performing structured processing on the first point cloud data to obtain a structured map;

[0006] Obtaining initial pose data of the subway driving, obtaining trajectory pose data by calculation based on the structured map and the initial pose data, and extracting initial clearance space corresponding to the trajectory pose data from the structured map;

[0007] Obtaining point cloud data of a real-time driving environment of the subway as second point cloud data, performing classification and correction on the second point cloud data, and extracting feature point cloud from each frame of image in the second point cloud data after classification and correction by calculation;

[0008] Obtaining inter-frame pose data of the subway by calculation based on the feature point cloud corresponding to adjacent two frames of image;

[0009] Obtaining real-time pose data of the subway by calculation based on the inter-frame pose data, the trajectory pose data and the structured map;

[0010] The real-time clearance space of the subway is obtained by calculating the offset of the real-time pose data compared with the trajectory pose data and the initial clearance space.

[0011] Based on the same inventive concept, the application further provides a subway clearance space detection device, comprising:

[0012] The structured map module is configured to obtain point cloud data of all target objects in a subway driving environment as first point cloud data, and perform structured processing on the first point cloud data to obtain a structured map.

[0013] The initial clearance space module is configured to obtain initial pose data of a subway driving, obtain trajectory pose data by calculating based on the structured map and the initial pose data, and extract an initial clearance space corresponding to the trajectory pose data from the structured map.

[0014] The feature point cloud module is configured to obtain point cloud data of a real-time driving environment of a subway as second point cloud data, perform classification and correction on the second point cloud data, and extract feature point clouds from each frame of image of the second point cloud data after classification and correction by calculation.

[0015] The inter-frame pose module is configured to obtain inter-frame pose data of a subway by calculating based on the feature point clouds corresponding to adjacent two frames of image.

[0016] The real-time pose module is configured to obtain real-time pose data of a subway by calculating based on the inter-frame pose data, the trajectory pose data and the structured map.

[0017] The real-time clearance space module is configured to obtain real-time clearance space of a subway by calculating the offset of the real-time pose data compared with the trajectory pose data and the initial clearance space.

[0018] Based on the same inventive concept, the application further provides an electronic device, comprising a memory, a processor and a computer program stored in the memory and executable by the processor, wherein the processor realizes the method as described above when executing the computer program.

[0019] Based on the same inventive concept, the application further provides a non-transitory computer readable storage medium, which stores computer instructions for causing a computer to execute the method as described above.

[0020] As described above, the subway clearance space detection method, device, equipment, and storage medium provided in this application, by pre-constructing a structured map, extracting the initial clearance space, and acquiring the subway's trajectory pose data, calculates the offset between the real-time pose data and the trajectory pose data during actual subway operation. Based on the initial clearance space, the real-time clearance space is then calculated. Using the method of pre-constructing a structured map and pre-extracting the initial clearance space, only the initial clearance space needs to be transformed when detecting the real-time clearance space, reducing the computational load, improving detection efficiency, real-time performance, and subway operational safety. Furthermore, the subway is equipped with a long-range sensing system that can determine whether distant obstacles have encroached on the clearance, further ensuring the operational safety of subway trains. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating the subway clearance space detection method according to an embodiment of this application.

[0023] Figure 2 This is a schematic diagram of a structured map according to an embodiment of this application;

[0024] Figure 3 This is a schematic diagram of the first planar boundary space according to an embodiment of this application;

[0025] Figure 4 This is a schematic diagram of the initial bounding space according to an embodiment of this application;

[0026] Figure 5 This is a schematic diagram illustrating the method for dividing ground point clouds and non-ground point clouds according to an embodiment of this application;

[0027] Figure 6 This is a schematic diagram of the subway clearance space detection device according to an embodiment of this application;

[0028] Figure 7 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0030] It should be noted that, unless otherwise defined, technical terms or scientific terms used in the embodiments of the present application shall have the common meaning understood by one of ordinary skill in the art to which the embodiments of the present application belong. The terms "first", "second", and similar terms used in the embodiments of the present application do not denote any order, quantity, or importance, but are merely used to distinguish different components. The terms "include", "contain", and similar terms mean that the elements or objects before the terms encompass the elements or objects listed after the terms and their equivalents, and do not exclude other elements or objects. The terms "connect" or "connected" and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The terms "upper", "lower", "left", "right", and the like are merely used to represent relative positional relationships, and when the absolute positions of the described objects change, the relative positional relationships can also change accordingly.

[0031] As described in the background, it is of great safety significance to detect the clearance space of the subway in real time and determine whether there is an intrusion risk during the operation of the subway. At present, for the detection of the clearance space of the subway, the scheme adopted is mostly to obtain the tunnel cross-section point cloud by a three-dimensional laser scanning detection device, convert it into a cross-section image, and then compare the cross-section image of the tunnel with the clearance map of the subway, and obtain the clearance analysis result according to the comparison result and the clearance space detection standard of the subway. The problems of the above scheme are as follows: the high-precision three-dimensional laser scanning detection device is high in cost, and multiple sets of high-precision three-dimensional laser scanning detection devices are needed for each subway, which is difficult to be used in large scale; the above scheme is mainly used for clearance space detection in the process of subway tunnel construction, and thus has low automation and low real-time performance, and the real-time position of the train and the clearance space in front of the train cannot be obtained in real time during the operation of the subway, so that it is impossible to ensure whether there is an obstacle in front of the train to cause intrusion hazard to the operation safety of the subway train.

[0032] The present application uses high-precision point cloud surveying equipment and inertial navigation devices to pre-collect high-precision point cloud data and trajectory pose data of the subway operating environment, processes to obtain a structured map and an initial clearance space, and then combines the real-time data of the laser radar and the inertial measurement unit (IMU) loaded on the subway train to match the structured map to obtain real-time pose data of the train, and calculates the real-time clearance space in front of the train based on the trajectory pose data and the initial clearance space, thereby effectively reducing the cost of loading the train, simplifying the calculation process of the real-time clearance space, improving the real-time performance, and ensuring the operation safety of the subway train. In addition, a long-distance perception system is also provided on the subway to detect long-distance obstacle targets and determine whether there is an intrusion risk, thereby further ensuring the safety of the subway operation.

[0033] Embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0034] The present application provides a metro clearance space detection method, referring to Figure 1 , comprising the following steps:

[0035] Step S101, obtaining point cloud data of all target objects in the metro driving environment as first point cloud data, and performing structured processing on the first point cloud data to obtain a structured map;

[0036] Specifically, in the present embodiment, a train loaded with a three-dimensional laser scanning detection device is used as a test vehicle, which runs at a constant speed of 40 km / h on the track, collects all point cloud data of the train driving environment as first point cloud data, and performs structured processing on the first point cloud data to generate a structured map. The test vehicle is slower than the actual running metro train, and the collection of point cloud data is more comprehensive and accurate. Moreover, only the three-dimensional laser scanning detection device needs to be loaded on the test vehicle to pre-construct the structured map, without the need to load it again on the actual running metro train, thereby reducing the loading cost and improving the economic benefit.

[0037] Step S102, obtaining initial pose data of the metro driving, obtaining trajectory pose data by calculation based on the structured map and the initial pose data, and extracting initial clearance space corresponding to the trajectory pose data from the structured map;

[0038] Specifically, in the present embodiment, the test vehicle is also loaded with an inertial navigation device, which collects position, speed and attitude angle data of the test vehicle as initial pose data. The initial pose data is fused with the structured map through an error Kalman (ESKF, Error State Kalman Filter) filter algorithm to obtain trajectory pose data of the test vehicle, and initial clearance space corresponding to each trajectory pose data is extracted from the structured map. Based on the structured map, the trajectory pose data of the test vehicle is pre-built and the corresponding initial clearance space is extracted, which provides a basis for later calculation of real-time pose data and real-time clearance space of the metro, improves the calculation efficiency, and thus improves the real-time performance. In some embodiments, in addition to using the ESKF filter algorithm, other optimization methods such as graph optimization can also be used to obtain the trajectory pose data of the test vehicle.

[0039] Step S103, obtaining point cloud data of the metro real-time driving environment as second point cloud data, classifying and correcting the second point cloud data, and extracting feature point cloud from each frame of image in the classified and corrected second point cloud data through calculation;

[0040] Specifically, in the embodiment, a mechanical rotating multi-line laser radar is loaded on the subway to collect point cloud data of the driving environment in real time as the second point cloud data in the rapid operation of the subway, the second point cloud data is classified and corrected, and the points in each frame of the second data point cloud after correction are calculated to extract the feature point cloud. Since the subway runs at a very high speed and there is frequent acceleration and deceleration when entering and leaving the station, all points in each frame of point cloud data collected by the laser radar are not collected at the same time, so it is necessary to correct the second point cloud data in combination with the pose change data of the subway to improve the accuracy of each frame of the second point cloud data. In some embodiments, in addition to using a mechanical rotating multi-line laser radar, other laser radars such as a semi-solid laser radar can also be used to collect point cloud data of the driving environment in real time as the second point cloud data.

[0041] In step S104, the inter-frame pose data of the subway is obtained by calculation based on the feature point clouds corresponding to the adjacent two frames of images.

[0042] Specifically, in the embodiment, the feature point cloud extracted from the second point cloud data of the current frame is registered with the feature point cloud extracted from the second point cloud data of the previous frame by the Iterative Closest Point (ICP) algorithm to obtain the inter-frame pose data of the subway. Compared with the large amount of second point cloud data, the number of points is reduced, the amount of calculation is greatly reduced, the calculation frequency is improved, and the real-time positioning requirement under the high-speed operation of the subway can be met.

[0043] In step S105, the real-time pose data of the subway is obtained by calculation based on the inter-frame pose data, the track pose data and the structured map.

[0044] In step S106, the offset of the real-time pose data compared with the track pose data is calculated, and the real-time bounding space of the subway is obtained by calculation based on the offset and the initial bounding space.

[0045] In some embodiments, the structured processing includes: performing a first preprocessing on the first point cloud data to eliminate noise points and moving target points in the first point cloud data; classifying the first point cloud data after the first preprocessing based on the target object category; segmenting the classified first point cloud data based on the line section and the subway station; establishing a first spatial index structure based on the segmented first point cloud data, and performing curvature calculation on the first point cloud data.

[0046] Specifically, for the first point cloud data collected by the three-dimensional laser scanning detection device, noise points and moving target points are removed therefrom; after the removal, the first point cloud data is classified based on semantic information of the point cloud data by a semi-automatic method, and the first point cloud data is given corresponding semantic information, including cable point cloud, pole point cloud, building point cloud, terrain point cloud, track point cloud, cantilever point cloud, vegetation point cloud, equipment point cloud, ballast point cloud, tunnel wall point cloud, etc.; based on the classification, the first point cloud data is segmented according to line sections and subway platforms; a kd-tree neighborhood query structure is established for the segmented first point cloud data, and the curvature of points contained in the classified and segmented first point cloud data is calculated to obtain a structured map as shown in Figure 2 .

[0047] In some embodiments, in addition to using the kd-tree data structure to save the point cloud data, other data structures such as OctoMap can also be used to save the point cloud data for convenient and fast searching of points in the point cloud that meet the conditions.

[0048] In some embodiments, the extracting of the initial clearance space corresponding to the trajectory pose data from the structured map comprises: all point cloud data in the structured map is recorded as a set S1; the trajectory pose data is recorded as a set S2; taking the mass center of the head of the subway before departure as a reference point, based on the profile parameters of the subway and the clearance space safety coefficient, a first plane clearance space at the head is obtained by calculation, the first plane clearance space is filled with point cloud, and point cloud data in the first plane clearance space is recorded as a set S3; based on the set S2 and the set S3, a plane clearance space corresponding to each of the trajectory pose data in the set S2 is obtained by calculation; all the plane clearance spaces form the initial clearance space.

[0049] Specifically, the direction in which the subway advances is taken as the z-axis, the upward direction is taken as the y-axis, and the rightward direction is taken as the x-axis to construct a point cloud map coordinate system;

[0050] All point cloud data contained in the structured map is recorded as a set S1;

[0051] The trajectory pose data includes path points and attitude angles, and taking the mass center of the head of the test vehicle as a reference point, the set of path points p={x, y, z} and attitude angles of the test vehicle during the collection of the first point cloud map is recorded as a set S2={(p0, q0), (p1, q1), … (p n , q n )};

[0052] With the mass center of the head of the test vehicle before departure as the reference point, set the path point of the test vehicle at this time as p0, and the attitude angle as q0. According to the profile parameters and the clearance space safety factor of the subway, the xoy first plane clearance space at the head at this time is calculated. The xoy first plane clearance space is filled with a minimum size grid of 1 cm. The center point coordinates of each grid are extracted. The set of all grid center points is denoted as the point cloud data in the first plane clearance space, denoted as set S3={(x0,y0,0),(x1,y1,0)…(x n ,y n ,0)}, as shown in the dashed box in Figure 3 , which is the first plane clearance space;

[0053] Let T n =f n (S2,S3), T n represents the corresponding plane clearance space at any trajectory pose; f n represents the corresponding conversion relationship of the first plane clearance space to the current path point and the attitude angle at any trajectory pose of the subway. Let the current path point p n correspond to the translation vector t of p0. The current attitude angle q n corresponds to the rotation matrix R of q0. Then the coordinates of the current plane clearance space are (x′ n ,y′ n ,z′ n ) T =R×(x0,y0,0) T +t;

[0054] The sum of all plane clearance spaces constitutes the initial clearance space, denoted as The initial clearance space corresponding to the trajectory pose data is obtained, as shown in the dashed box in Figure 4 , which is the initial clearance space corresponding to the trajectory pose data. The correspondence between the structured map and the trajectory pose data and the initial clearance space is written into a file for real-time positioning and real-time clearance space calculation.

[0055] In some embodiments, based on the kd-tree neighborhood query structure established based on the first point cloud data, it is determined whether the points in set S4 interfere with the points in set S1 or the distance between any two points is less than a threshold value. In response to the points in the plane clearance space T n interfering with the points in set S1 and the distance between any two points being less than 30 cm, the cross-sectional profile of the test vehicle at the point (p n ,q n ) is manually measured. The point cloud map is corrected, and then the corrected initial clearance space is extracted.

[0056] In some embodiments, classifying and correcting the second point cloud data, and extracting feature point clouds from each frame of the classified and corrected second point cloud data by calculation, includes: performing a second preprocessing on the second point cloud data to remove invalid points; dividing the second point cloud data after the second preprocessing into ground point clouds and non-ground point clouds, and clustering the non-ground point clouds to obtain clustered point clouds; acquiring the real-time acceleration, real-time angular velocity, and real-time Euler angles of the subway during operation, and correcting the ground point clouds and clustered point clouds based on the real-time acceleration, real-time acceleration, and real-time Euler angles to obtain corrected ground point clouds and corrected clustered point clouds; performing curvature calculation on the corrected ground point clouds, and obtaining planar feature point clouds based on the curvature calculation results; performing curvature calculation on the corrected clustered point clouds, and obtaining polyline feature point clouds based on the curvature calculation results; and using the planar feature point clouds and the polyline feature point clouds as the feature point clouds.

[0057] Specifically, the second point cloud data collected by the lidar is preprocessed to remove invalid points and points outside the lidar detection angle range;

[0058] By calculating the coordinates of adjacent line points of the lidar and the angle between them and the lidar mounting plane, it is determined whether the ground point cloud threshold condition is met, thus dividing the second point cloud data into ground point cloud and non-ground point cloud. In this embodiment, as... Figure 5 As shown, assuming the lidar is installed horizontally parallel to the ground, it emits two adjacent beams, A1 and A2, at the same rotation angle and at the same time. A1 falls on ground point B1, and A2 falls on non-ground point B2. If the angle α between the line connecting points B1 and B2 and the ground is greater than 1 degree, then a non-ground point is determined to exist; if the angle α between the line connecting points B1 and B2 and the ground is less than or equal to 1 degree, then both are ground points. Based on this, the second point cloud data is divided into ground point cloud data and non-ground point cloud data. A breadth-first search (BFS) clustering method is used on the non-ground point cloud data, setting the minimum cluster point and clustering plane conditions to obtain clustered and non-clustered point clouds.

[0059] Because the subway runs at a very high speed, the point cloud data acquired by the lidar is not collected at the same time. In this embodiment, the lidar collects data from two adjacent points at a time difference of 100ms. The subway is equipped with an IMU to collect the acceleration, angular velocity and Euler angle data of the subway within this 100ms. Based on the collected subway pose change data, the ground point cloud data and clustered point cloud data are corrected to remove distortion and obtain corrected ground point cloud data and corrected clustered point cloud data.

[0060] The square sum of the spatial distances of the five points before and after the different scanning moments of the points in the same beam is used to calculate the curvatures of the points in the corrected ground point cloud and the corrected clustered point cloud, 20 points with the minimum curvatures in the ground point cloud are extracted as the plane feature point cloud, and 20 points with the maximum curvatures in the clustered point cloud are extracted as the polyline feature point cloud, and the plane feature point cloud and the polyline feature point cloud form the feature point cloud.

[0061] In some embodiments, the real-time pose data of the subway is calculated based on the inter-frame pose data, the trajectory pose data and the structured map, including: taking the feature point cloud corresponding to the inter-frame pose data as a first feature point cloud; extracting the trajectory pose data closest to the inter-frame pose data from the trajectory pose data as a first trajectory pose data, and extracting the point cloud data of the previous and subsequent frames corresponding to the first trajectory pose data from the structured map; extracting feature point clouds from the point cloud data of each frame in the previous and subsequent frames as second feature point clouds; and obtaining the real-time pose data of the subway based on the first feature point cloud and the second feature point cloud through a nonlinear optimization algorithm.

[0062] Specifically, the real-time second point cloud data of the subway at the time of the inter-frame pose data is obtained, and the plane feature point cloud and the polyline feature point cloud in the second point cloud data at this time are extracted as the first feature point cloud according to the above method;

[0063] According to the inter-frame pose data (p m ,q m ), the closest pose data is extracted from the trajectory pose data as a first trajectory pose data (p m1 ,q m1 ), and the point cloud data of the previous and subsequent five frames corresponding to the first trajectory pose data (p m1 ,q m1 ) is extracted from the structured map, and the plane feature point cloud and the polyline feature point cloud are extracted from the point cloud data of each frame of the previous and subsequent five frames as second feature point clouds according to the above method, and there are 10 groups of second feature point clouds at this time;

[0064] The 10 groups of second feature point clouds are respectively registered with the first feature point cloud through the ICP algorithm to obtain 10 pose data, and the real-time pose data is finally optimized through the nonlinear optimization method of the least square method based on the 10 pose data.

[0065] The accuracy of the inter-frame pose data is low, and the error will be amplified over time. Re-matching and positioning with the trajectory pose data and the structured map can make the error converge. The output frequency of the real-time pose of the subway train is ensured by calculating the inter-frame pose data, and the gradual convergence of the error is ensured by re-calculating the real-time pose data based on the inter-frame pose data, the trajectory pose data and the structured map.

[0066] In some embodiments, calculating the offset of the real-time pose data relative to the trajectory pose data includes: extracting the data closest to the real-time pose data from the trajectory pose data as second trajectory pose data, and calculating the offset of the real-time pose data relative to the second trajectory pose data, wherein the offset includes a rotation matrix and a translation vector.

[0067] Specifically, based on real-time pose data (p w ,q w Extract the closest pose data from the trajectory pose data as the second trajectory pose data (p) w1 ,q w1 ), calculate real-time pose data (p w ,q w Compared to the second trajectory pose data (p) w1 ,q w1 The rotation matrix K and translation vector d of )

[0068] Based on the real-time pose data of the subway and the initial bounded space pre-extracted from the structured map, the real-time bounded space during subway operation can be calculated. Second trajectory pose data (p) is then obtained. w1 ,q w1 The corresponding initial bounded space T w1 Let T be at this time. w1 The point cloud set contained is S w1 , for S w1 All points in the coordinate system can be transformed using the rotation matrix K and translation vector d to obtain the real-time bounded space H. w At this point, extracting the real-time bounded space only requires calculating the coordinate transformation of the initial bounded space point cloud extracted from the structured map, greatly reducing the computational load and improving detection accuracy. Let H be... w The point cloud set contained is S h Based on the kd-tree neighborhood query structure built from the first point cloud data, the judgment set S is determined. h Whether a point in set S1 interferes with a point in set S1 or whether the distance between any two points is less than a threshold.

[0069] In some embodiments, long-distance point cloud data in the subway driving environment is acquired as third point cloud data, obstacle point cloud data is acquired based on the third point cloud data and the structured map, and in response to determining that the obstacle point cloud data and the real-time boundary space have boundary intrusion and the degree of boundary intrusion is greater than a preset safety threshold, the subway deceleration or stopping is performed.

[0070] Specifically, a long-range sensing system is installed on the subway, including a laser radar and a long-focus camera. The laser radar is used to collect third point cloud data in front of the subway running, and the long-focus camera is used to assist the laser radar to obtain semantic information of the third point cloud data. The invalid points in the third point cloud data are removed, and the ground point cloud and the tunnel wall point cloud in the third point cloud data are divided. The remaining point cloud data in the third point cloud data is clustered to obtain a clustering target cluster. The judgment condition of the clustering target can be appropriately relaxed, so as to reduce the false negative rate of the obstacle. The first point cloud data in front of the current pose is extracted from the structured map, and based on the semantic information of the first point cloud data and the semantic information of the third point cloud data in the clustering target cluster, the point cloud data in the clustering target cluster is filtered to remove the same point cloud data, and the remaining point cloud data is recorded as the obstacle point cloud data. Based on the obstacle point cloud data, a kd-tree neighborhood query structure is constructed, and it is judged whether the point cloud data S h whether there is an invasion boundary intersection with the obstacle point cloud data. In response to determining that the real-time limit space and the obstacle point cloud data have an invasion boundary intersection, and the nearest distance between the obstacle point and the real-time limit space is less than or equal to 40 cm, the operation of subway deceleration or parking is performed.

[0071] In some embodiments, in addition to using a kd-tree data structure to save point cloud data, other data structures such as OctoMap, MVOG, etc. can also be used to save point cloud data for convenient and fast searching of points in the point cloud that meet the conditions.

[0072] It should be noted that the method of the embodiments of the present application can be executed by a single device, such as a computer or a server, etc. The method of the embodiments of the present application can also be applied to a distributed scenario, and completed by multiple devices cooperating with each other. In this distributed scenario, one of the multiple devices can only execute one or more steps in the method of the embodiments of the present application, and the multiple devices can interact with each other to complete the method.

[0073] It should be noted that some embodiments of the present application have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order than the order described above and still achieve desirable results. Additionally, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In some implementations, multitasking and parallel processing can be advantageous.

[0074] Based on the same inventive concept, the present application also provides a subway limit space detection device corresponding to the method of any of the above embodiments.

[0075] Reference Figure 6The metro gauge space detection device comprises:

[0076] The structured map module 601 is configured to acquire point cloud data of all target objects in a metro driving environment as first point cloud data, and perform structured processing on the first point cloud data to obtain a structured map.

[0077] The initial gauge space module 602 is configured to acquire initial pose data of metro driving, obtain trajectory pose data by calculation based on the structured map and the initial pose data, and extract initial gauge space corresponding to the trajectory pose data from the structured map.

[0078] The feature point cloud module 603 is configured to acquire point cloud data of a real-time driving environment of the metro as second point cloud data, perform classification and correction on the second point cloud data, and extract feature point cloud from each frame of image in the classified and corrected second point cloud data by calculation.

[0079] The inter-frame pose module 604 is configured to obtain inter-frame pose data of the metro by calculation based on the feature point cloud corresponding to adjacent two frames of image.

[0080] The real-time pose module 605 is configured to obtain real-time pose data of the metro by calculation based on the inter-frame pose data, the trajectory pose data and the structured map.

[0081] The real-time gauge space module 606 is configured to calculate an offset of the real-time pose data compared with the trajectory pose data, and obtain real-time gauge space of the metro by calculation based on the offset and the initial gauge space.

[0082] For the convenience of description, the above device is described in various modules in terms of functions. Of course, the functions of the modules can be implemented in one or more software and / or hardware in the implementation of the present application.

[0083] The device of the above embodiment is used to implement the corresponding metro gauge space detection method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which are not described here.

[0084] Based on the same inventive concept, the present application also provides an electronic device corresponding to the above-mentioned any embodiment method, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the metro gauge space detection method of any one of the above embodiments.

[0085] Corresponding to the method of any of the above embodiments based on the same inventive concept, the application further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the metro gauge space detection method of any of the above embodiments when executing the program.

[0086] Figure 7 A more specific hardware structure of an electronic device provided by the embodiment is shown, which can include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 for internal communication within the device.

[0087] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits, etc., for executing related programs to implement the technical solutions provided by the embodiments of the present specification.

[0088] The memory 1020 can be implemented by a ROM (Read Only Memory), a RAM (Random Access Memory), a static storage device, a dynamic storage device, etc. The memory 1020 can store an operating system and other application programs, and when the technical solutions provided by the embodiments of the present specification are implemented by software or firmware, the related program codes are stored in the memory 1020 and called and executed by the processor 1010.

[0089] The input / output interface 1030 is used to connect input / output modules to realize information input and output. The input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. The input device can include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device can include a display, a speaker, a vibrator, an indicator light, etc.

[0090] The communication interface 1040 is used to connect a communication module (not shown in the figure) to realize the communication interaction between the device and other devices. The communication module can realize communication through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.).

[0091] Bus 1050 includes a path for transferring information between the various components (e.g., processor 1010, memory 1020, input / output interface 1030, and communication interface 1040) of the device.

[0092] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device can also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device can also only contain the components necessary to implement the embodiments of the present application, and does not have to contain all the components shown in the figure.

[0093] The electronic device of the above embodiment is used to implement the corresponding metro gauge space detection method in any of the preceding embodiments, and has the beneficial effects of the corresponding method embodiments, which are not repeated here.

[0094] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present application also provides a non-transitory computer readable storage medium, which stores computer instructions for causing the computer to execute the metro gauge space detection method according to any of the above embodiments.

[0095] The computer readable medium of the present embodiment includes permanent and non-permanent, removable and non-removable media, which can be realized by any method or technology to store information. The information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape magnetic disk storage or other magnetic storage device, or any other non-transmission medium that can be used to store information accessible by a computing device.

[0096] The computer instructions stored in the storage medium of the above embodiment are used to cause the computer to execute the metro gauge space detection method according to any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which are not repeated here.

[0097] Those of ordinary skill in the art will realize that the foregoing discussion of any of the embodiments has been presented for the purpose of illustration and description and is not intended to be exhaustive or to limit the application to the precise forms described, and that various adaptations and modifications are possible within the scope and spirit of the application. For example, while the embodiments discussed above have been described in the context of a memory device, other memory architectures (e.g., dynamic RAM (DRAM)) can use the embodiments discussed.

[0098] In addition, to simplify the description and discussion, and so as not to make the embodiments of the application difficult to understand, the well-known power / ground connections to integrated circuit (IC) chips and other components can or can not be shown in the provided drawings. Further, devices can be shown in block diagram form so as not to make the embodiments of the application difficult to understand, and this also takes into account the fact that the details regarding the implementation of these block diagram devices are highly dependent on the platform in which the embodiments of the application are to be implemented (i.e., these details should be well within the understanding of one of ordinary skill in the art). Where specific details (e.g., circuitry) are set forth in order to describe an illustrative embodiment of the application, it should be apparent to one of ordinary skill in the art that the embodiments of the application can be practiced without or with variations of these specific details. Thus, the description should not be considered to be limiting in nature.

[0099] While the application has been described in connection with specific embodiments thereof, it will be understood that many modifications, variations and alternatives will be apparent to those skilled in the art as a result of the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) can use the embodiments discussed.

[0100] The embodiments of the application are intended to cover all such modifications, variations and alternatives as come within the scope of the appended claims. Thus, any and all such modifications, variations and alternatives are intended to be included within the scope of the application.

Claims

1. A method for detecting confined spaces in subway systems, characterized in that, include: The point cloud data of all target objects in the subway driving environment is obtained as the first point cloud data, and the first point cloud data is processed in a structured manner to obtain a structured map. The initial pose data of the subway is obtained, and the trajectory pose data is calculated based on the structured map and the initial pose data. The initial bound space corresponding to the trajectory pose data is extracted from the structured map. The point cloud data of the real-time subway driving environment is acquired as the second point cloud data. The second point cloud data is classified and corrected. Feature point cloud is extracted from each frame of the classified and corrected second point cloud data by calculation. Based on the feature point cloud corresponding to two adjacent frames, the inter-frame pose data of the subway is calculated. Based on the inter-frame pose data, the trajectory pose data, and the structured map, the real-time pose data of the subway is calculated. The offset of the real-time pose data relative to the trajectory pose data is calculated, and the real-time boundary space of the subway is obtained based on the offset and the initial boundary space.

2. The subway clearance space detection method according to claim 1, characterized in that, The structuring process includes: The first point cloud data is preprocessed to remove noise points and moving target points from the first point cloud data; The first point cloud data after the first preprocessing is classified based on the target object category; The first point cloud data after classification is segmented based on the line section and subway platform; A first spatial index structure is established based on the segmented first point cloud data, and curvature calculation is performed on the first point cloud data.

3. The subway clearance space detection method according to claim 1, characterized in that, The step of extracting the initial bounded space corresponding to the trajectory pose data from the structured map includes: The entire point cloud data in the structured map is denoted as set S1; The trajectory pose data is denoted as set S2; Using the centroid of the train head before departure as the reference point, based on the contour parameters of the subway and the safety factor of the clearance space, the first planar clearance space at the train head is calculated, the first planar clearance space is filled with point cloud, and the point cloud data in the first planar clearance space is denoted as set S3. Based on sets S2 and S3, the planar bounded space corresponding to each trajectory pose data in set S2 is obtained by calculation. All the planar bounded spaces constitute the initial bounded space.

4. The subway clearance space detection method according to claim 1, characterized in that, The classification and correction of the second point cloud data, by calculating and extracting feature point clouds from each frame of the classified and corrected second point cloud data, includes: The second point cloud data is preprocessed a second time to remove invalid points from the second point cloud data; The second point cloud data after the second preprocessing is divided into ground point cloud and non-ground point cloud, and the non-ground point cloud is clustered to obtain clustered point cloud; The real-time acceleration, real-time angular velocity, and real-time Euler angles of the subway during operation are obtained. Based on the real-time acceleration, real-time angular velocity, and real-time Euler angles, the ground point cloud and the clustered point cloud are corrected to obtain the corrected ground point cloud and the corrected clustered point cloud. The curvature of the corrected ground point cloud is calculated, and a planar feature point cloud is obtained based on the curvature calculation results; Curvature calculation is performed on the corrected clustered point cloud, and a polyline feature point cloud is obtained based on the curvature calculation result; The planar feature point cloud and the polyline feature point cloud are used as the feature point cloud.

5. The subway clearance space detection method according to claim 1, characterized in that, The process of calculating real-time pose data for the subway based on the inter-frame pose data, the trajectory pose data, and the structured map includes: The feature point cloud corresponding to the inter-frame pose data is used as the first feature point cloud; The trajectory pose data closest to the inter-frame pose data is extracted from the trajectory pose data as the first trajectory pose data. Point cloud data of multiple frames before and after the current frame corresponding to the first trajectory pose data are extracted from the structured map. Feature point cloud is extracted from the point cloud data of each of the multiple frames before and after the first trajectory pose data as the second feature point cloud. The real-time pose data of the subway is obtained through a nonlinear optimization algorithm based on the first feature point cloud and the second feature point cloud.

6. The subway clearance space detection method according to claim 1, characterized in that, The calculation of the offset of the real-time pose data relative to the trajectory pose data includes: Extract the data that is closest to the real-time pose data from the trajectory pose data as the second trajectory pose data, and calculate the offset of the real-time pose data relative to the second trajectory pose data. The offset includes the rotation matrix and the translation vector.

7. The subway clearance space detection method according to claim 1, characterized in that, Also includes: Long-distance point cloud data in the subway operating environment is acquired as third point cloud data. Obstacle point cloud data is acquired based on the third point cloud data and the structured map. In response to determining that the obstacle point cloud data and the real-time boundary space have boundary intrusion and the degree of boundary intrusion is greater than a preset safety threshold, the subway deceleration or stopping is executed.

8. A subway clearance space detection device, characterized in that, include: The structured map module is configured to acquire point cloud data of all target objects in the subway driving environment as first point cloud data, and perform structured processing on the first point cloud data to obtain a structured map. The initial bound space module is configured to acquire the initial pose data of the subway, calculate the trajectory pose data based on the structured map and the initial pose data, and extract the initial bound space corresponding to the trajectory pose data from the structured map. The feature point cloud module is configured to acquire point cloud data of the real-time subway driving environment as the second point cloud data, classify and correct the second point cloud data, and extract feature point cloud from each frame of the classified and corrected second point cloud data through calculation. The inter-frame pose module is configured to calculate the inter-frame pose data of the subway based on the feature point cloud corresponding to two adjacent frames. The real-time pose module is configured to calculate the real-time pose data of the subway based on the inter-frame pose data, the trajectory pose data, and the structured map. The real-time clearance space module is configured to calculate the offset of the real-time pose data relative to the trajectory pose data, and to calculate the real-time clearance space of the subway based on the offset and the initial clearance space.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing the computer to perform the method of any one of claims 1 to 7.

Images