Railway communication device interested area determination method, device, equipment and medium
By using time-of-flight sensors to acquire point cloud data in the communication cabinet, the communication equipment area across the field of view can be identified and aggregated, solving the problem that visual sensors cannot fully cover the area. This enables efficient and accurate equipment identification and reduces maintenance costs.
Patent Information
- Application Number
- CN202511161351.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-19
- Publication Date
- 2025-10-21
- Estimated Expiration
- 2045-08-19
Smart Images
Figure CN120655905B_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of computer vision technology, and in particular to a method, apparatus, device, and medium for determining a region of interest for railway communication equipment. Background Art
[0002] With the rapid development of communication network infrastructure, communication cabinets, as important network nodes, are widely deployed in various fields such as 5th Generation Mobile Communication Technology (5G) base stations, railway communications, power communications, and data centers.
[0003] Related technologies deploy visual sensors within communication cabinets, coupled with image recognition algorithms, to enable real-time remote monitoring of communication equipment within the cabinets. However, due to the limitations of the visual sensor's field of view, a single field of view may not fully cover the entire surface of a communication device. The same communication device may appear in multiple adjacent fields of view, forming multiple independent point cloud clusters. Accurately identifying the area containing communication devices across fields of view has become a pressing issue. Summary of the Invention
[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the present disclosure provides a method, apparatus, device and medium for determining an area of interest of railway communication equipment.
[0005] The present disclosure provides a method for determining an area of interest of a railway communication device, including:
[0006] Acquire multiple candidate point cloud data generated by performing regional measurement on the communication cabinet, and determine multiple measurement area pairs based on the multiple candidate point cloud data; wherein the communication cabinet has multiple communication devices arranged therein, the candidate point cloud data corresponds one-to-one to the measurement areas, and the measurement area pairs include two adjacent measurement areas in which the same communication device is located;
[0007] Determining a plurality of connected boundaries corresponding to the plurality of measurement area pairs; wherein the connected boundary is a region boundary between two measurement areas included in the measurement area pair;
[0008] Acquire a plurality of boundary point cloud data groups generated by scanning and measuring the plurality of connected boundaries, and determine the connected boundaries corresponding to the boundary point cloud data groups that meet the surface continuity condition as target boundaries;
[0009] Grouping the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups;
[0010] A region of interest of the communication device is determined based on the surface point cloud cluster group.
[0011] The present disclosure also provides a device for determining an area of interest of a railway communication device, including:
[0012] an area pair determination module, configured to obtain a plurality of candidate point cloud data generated by performing area-by-area measurement on a communication cabinet, and determine a plurality of measurement area pairs based on the plurality of candidate point cloud data; wherein the communication cabinet contains a plurality of communication devices, the candidate point cloud data corresponds one-to-one to the measurement areas, and the measurement area pairs include two adjacent measurement areas containing the same communication device;
[0013] A first boundary determination module is configured to determine a plurality of connected boundaries corresponding to the plurality of measurement area pairs; wherein the connected boundary is a region boundary between two measurement areas included in the measurement area pair;
[0014] a second boundary determination module, configured to obtain a plurality of boundary point cloud data groups generated by scanning and measuring the plurality of connected boundaries, and determine the connected boundaries corresponding to the boundary point cloud data groups that meet the surface continuity condition as target boundaries;
[0015] a grouping module, configured to group the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups;
[0016] The region determination module is configured to determine a region of interest of the communication device according to the surface point cloud cluster group.
[0017] An embodiment of the present disclosure also provides an electronic device, which includes: a processor; a memory for storing executable instructions of the processor; the processor is used to read the executable instructions from the memory and execute the instructions to implement the method for determining the area of interest of the railway communication equipment provided in the embodiment of the present disclosure.
[0018] An embodiment of the present disclosure further provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program is used to execute the method for determining an area of interest of a railway communication device as provided in the embodiment of the present disclosure.
[0019] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has the following advantages: a solution for determining the region of interest of railway communication equipment provided in the embodiments of the present disclosure, the method comprising: obtaining a plurality of candidate point cloud data generated by performing regional measurement on a communication cabinet, and determining a plurality of measurement region pairs based on the plurality of candidate point cloud data; wherein a plurality of communication devices are arranged in the communication cabinet, the candidate point cloud data correspond one-to-one to the measurement region, and the measurement region pair includes two adjacent measurement regions where the same communication device exists; determining a plurality of connected boundaries corresponding to the plurality of measurement region pairs; wherein the connected boundary is the region boundary between the two measurement regions included in the measurement region pair; obtaining a plurality of boundary point cloud data groups generated by scanning and measuring the plurality of connected boundaries, and determining the connected boundaries corresponding to the boundary point cloud data groups that meet the surface continuity condition as target boundaries; grouping and processing the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundaries to obtain a plurality of surface point cloud cluster groups; and determining the region of interest of the communication equipment based on the surface point cloud cluster groups.
[0020] By adopting the above technical solution, candidate point cloud data is obtained by grid scanning the communication cabinet, and adjacent pairwise measurement areas containing the same communication equipment are determined. The connected boundary between the two paired measurement areas is scanned and measured to obtain multiple boundary point cloud data. When multiple boundary point cloud data groups meet the plane continuity condition, the connected boundary is determined as the target boundary, and point cloud clusters are aggregated based on the target boundary and the area of interest of the communication equipment is determined. By scanning the connected boundary that is potentially the splicing boundary of the point cloud cluster and determining that the scanned boundary point cloud data meets the plane continuity condition, it is indicated that the connected boundary is actually the boundary of the area crossed by the communication equipment. Then, the point cloud clusters across the field of view are aggregated and the area of interest is identified based on the connected boundary, thereby realizing the accurate and efficient determination of the area of interest of the communication equipment when the same communication equipment appears in multiple adjacent fields of view. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.
[0022] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0023] Figure 1 A schematic flow chart of a method for determining an area of interest of railway communication equipment provided in an embodiment of the present disclosure;
[0024] Figure 2 A schematic diagram of an installation of a time-of-flight sensor provided in an embodiment of the present disclosure;
[0025] Figure 3 A schematic diagram of a measurement position provided in an embodiment of the present disclosure;
[0026] Figure 4 A schematic flow chart of another method for determining an area of interest for railway communication equipment provided in an embodiment of the present disclosure;
[0027] Figure 5 A schematic plan view of a communication cabinet provided in an embodiment of the present disclosure;
[0028] Figure 6 A schematic structural diagram of a device for determining a region of interest for railway communication equipment provided by an embodiment of the present disclosure;
[0029] Figure 7 A schematic structural diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0030] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.
[0031] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.
[0032] With the rapid development of communications network infrastructure, communications cabinets, as critical network nodes, are widely deployed in various fields, including 5G base stations, railway communications, power communications, and data centers. Inspection of communications equipment in these cabinets primarily relies on manual on-site inspections, which presents the following technical challenges: Operation and maintenance costs continue to rise. With the massive growth in the number of communications cabinets, manual on-site inspections require a large number of technicians to regularly visit various sites, resulting in significant labor costs. Real-time performance is poor, with manual inspections typically conducted on a weekly or monthly basis, failing to provide real-time monitoring. Status changes and faults in critical communications equipment are often delayed, impacting network service continuity. Inspection standards are inconsistent, and manual visual inspections are susceptible to subjective factors such as operator experience and fatigue. Accuracy and consistency are difficult to ensure, especially in densely packed cabinet environments. Inspection efficiency is a significant constraint, with manual inspections of a single cabinet typically taking 2-4 hours. With the large-scale construction of infrastructure such as railways and 5G, traditional inspection methods are no longer sufficient for rapid deployment and efficient operations and maintenance.
[0033] To address the aforementioned issues with manual inspections, online remote automatic monitoring technology based on fixed visible light cameras has been promoted. By deploying cameras within communications cabinets and integrating them with image recognition algorithms, remote, real-time monitoring of device status is achieved. However, this remote, real-time monitoring approach suffers from several issues: Regions of Interest (ROIs) rely on manual calibration. During the remote, real-time monitoring process, technicians must manually draw ROIs for each device within the cabinet. This time-consuming and labor-intensive process is a major obstacle to system deployment. ROI calibration is a significant workload. A single communications cabinet typically contains 10-30 communications devices, and manually calibrating each ROI takes 2-4 hours. For large-scale deployments of tens of thousands of sites, the calibration workload is enormous. Calibration accuracy depends on user experience, and ROIs calibrated by different technicians can vary significantly, resulting in inconsistent boundary accuracy, which directly impacts the performance and reliability of the subsequent automatic recognition algorithm. Maintenance and update costs are high. Whenever communications equipment is relocated, replaced, or newly added, personnel must be re-dispatched to perform on-site ROI calibration, resulting in high maintenance costs.
[0034] When automatically determining areas of interest, there may be a mismatch between the visual sensor's field of view and the size of the communications equipment. Specifically, the equipment within a communications cabinet is installed inside a standard cabinet. The surface width of the communications equipment is typically over 40 cm, while the effective field of view of the visual sensor is approximately 30 cm x 40 cm. Therefore, a single field of view cannot fully cover the width of the device surface. Two field of view units are required to fully cover the horizontal range of a device, resulting in horizontal coverage issues. Furthermore, depending on the height of the equipment, multiple field of view units may be required for vertical splicing, resulting in vertical coverage issues.
[0035] Since a single field of view cannot completely cover the surface of a communication device, the same communication device will appear in multiple adjacent fields of view during the grid scanning process, forming multiple independent point cloud clusters. How to accurately identify whether these cross-field point cloud clusters belong to the same device surface and perform correct surface splicing has become a technical problem that needs to be solved urgently. For point cloud clusters near the boundary of the field of view, it is necessary to determine whether they belong to the same device surface as the point cloud clusters in the adjacent field of view. In addition, the relevant technology mainly focuses on target recognition within a single field of view, and there is still a lack of solutions for splicing the surfaces of communication devices across fields of view in communication cabinet scenarios.
[0036] In order to solve the above problems, an embodiment of the present disclosure provides a method for determining an area of interest of railway communication equipment, which is described below in conjunction with specific embodiments.
[0037] Figure 1This is a flow chart of a method for determining an area of interest for railway communication equipment provided by an embodiment of the present disclosure. The method for determining an area of interest for railway communication equipment can be applied to an apparatus for determining an area of interest for railway communication equipment. The apparatus for determining an area of interest for railway communication equipment can be implemented using software and / or hardware, and can generally be integrated into an electronic device. Figure 1 As shown, the method for determining the region of interest of the railway communication equipment includes:
[0038] Step 101: Acquire multiple candidate point cloud data generated by performing regional measurement on a communication cabinet, and determine multiple measurement area pairs based on the multiple candidate point cloud data; wherein, multiple communication devices are arranged in the communication cabinet, the candidate point cloud data correspond one-to-one to the measurement areas, and the measurement area pair includes two adjacent measurement areas where the same communication device exists.
[0039] Among them, the communication cabinet, also known as the communication box room, can be a standardized cabinet for installing communication equipment. The regional measurement can be to divide the operation surface of the communication cabinet into a plurality of measurement areas by gridding, and then measure each measurement area. The operation surface can be the surface of the communication cabinet where manual operation is performed on the communication cabinet, and the operation surface can display one or more components of the switch, interface, and power supply of the communication equipment. The candidate point cloud data can be the point cloud data obtained by measuring part of the communication cabinet in the measurement area. This embodiment does not limit the measurement range corresponding to the candidate point cloud data. For example, the size of the measurement range of the candidate point cloud data can be consistent with the size of the measurement area, or the size of the measurement range of the candidate point cloud data can be smaller than the size of the measurement area.
[0040] The communication device may be a device used to build a communication network. This embodiment does not limit the type of the communication device. For example, the communication device may include one or more of a baseband unit (BBU), a remote radio unit (RRU), a transmission unit, a power supply unit, a patch panel, and an environmental monitoring unit. A measurement area pair may be used to record two adjacent measurement areas spanned by a communication device. The two measurement areas within a measurement area pair may be adjacent to each other horizontally or vertically, without limitation in this embodiment.
[0041] In some embodiments of the present disclosure, candidate point cloud data is collected by a time of flight (ToF) sensor, which is fixed on a mobile truss. The moving plane of the mobile truss is arranged relative to the operating surface of the communication cabinet. The mobile truss is used to move the time of flight sensor to a preset measurement position, and the measurement position is arranged in a two-dimensional array.
[0042] This embodiment does not limit the model of the time-of-flight sensor. For example, the time-of-flight sensor can be an industrial-grade time-of-flight camera. Specifically designed for use in darkroom environments within communication cabinets, the time-of-flight camera can have a depth detection range of 10-150 cm and a depth accuracy of ±1 mm. Furthermore, the camera is equipped with an infrared emitter, enabling an operating distance of over 1 meter in dark environments. The mobile truss can have two translational degrees of freedom to drive the movement of the time-of-flight sensor.
[0043] For example, Figure 2 A schematic diagram of the installation of a time-of-flight sensor provided in an embodiment of the present disclosure is shown in FIG. Figure 2 As shown, cabinet 201 can be a communications cabinet. The solid rectangular block within the cabinet can represent communications equipment. Frame 202 can be a door frame. Sensor 203 can be a time-of-flight sensor. The time-of-flight sensor can be mounted on a mobile truss, which can be a dual-axis precision truss and can be mounted on the cabinet's door frame. The two translational degrees of freedom correspond to the X-axis and Y-axis, respectively. The X-axis and Y-axis travels can be configured based on the cabinet door frame dimensions (for example, typically 1500mm-2500mm). The positioning accuracy of the mobile truss can be ±0.1mm. The mobile truss supports quick assembly and disassembly, facilitating its movement between different cabinets. The movement of the mobile truss can be controlled by a programmable logic controller (PLC) for efficient coordination. The PLC can serve as an execution unit, and the corresponding control unit can be a computer, which transmits the image data in real time to a remote monitoring system via a network. The time-of-flight sensor can transmit point cloud data to the control unit, which can perform image processing, algorithm analysis, and / or path planning based on the point cloud data. The control unit sends the control instruction to the execution unit, and the execution unit performs motion control, position feedback and / or safety monitoring on the mobile truss based on the control instruction.
[0044] The mobile truss can be installed on a mounting bracket that is adapted to the door frame size of a standard communication cabinet, supports the stable installation and precise positioning of the mobile truss, and has an adjustable installation height. Optionally, a lighting system and / or a network communication module can also be provided on the mobile truss. The lighting system can facilitate manual inspection and maintenance; the network communication module can support remote control and data transmission. This embodiment does not limit the configuration of the control unit. For example, the control unit can be configured with a processor, memory, and a graphics card, and the control unit can support real-time point cloud processing. This embodiment also does not limit the model of the execution unit.
[0045] The motion plane may be the plane in which the time-of-flight sensor moves, and the motion plane may be the plane defined by the two translational degrees of freedom of the mobile truss. The measurement position may be the position of the time-of-flight sensor during measurement. The measurement positions may be arranged in an array with equal spacing on the motion plane. Figure 3 A schematic diagram of a measurement position provided by an embodiment of the present disclosure, such as Figure 3 As shown, cabinet 301 may be a communication cabinet; the rectangle in the communication cabinet represents a communication device, and device 302 may be one of multiple communication devices; the black dot in the communication cabinet represents a measurement position, and position 303 may be one of multiple measurement positions. The measurement positions are arranged in an array.
[0046] In this embodiment, a mobile truss is deployed on the doorframe within a communications cabinet. This truss carries a time-of-flight sensor. This sensor replaces manual calibration of regions of interest (ROIs), enabling cross-field device surface stitching and high-precision automatic calibration of ROIs. A ROI determination device for railway communications equipment controls the mobile truss to move the time-of-flight sensor along a pre-set grid trajectory, collecting candidate point cloud data at each measurement location. Optionally, after determining the candidate point cloud data, it can undergo pre-processing, such as noise reduction.
[0047] In related technologies, automated calibration of regions of interest can be achieved based on visible light images. However, in the closed environment of a communications cabinet, visible light cameras rely on external lighting or fill light. Changes in lighting conditions can significantly affect detection accuracy, and errors are prone to occur in device boundary identification. It can be achieved based on lidar. Although the accuracy is high, the equipment cost is expensive, and multiple reflection interference is easily generated in the dense equipment environment of the cabinet, affecting the detection effect. It can be achieved based on structured light depth detection. However, mirror reflections are easily generated on the surface of the metal cabinet, affecting the recognition of structured light patterns. It is also sensitive to ambient light and lacks stability. It can be achieved based on binocular stereo vision, but a complex binocular positioning process is required, and it is highly dependent on the texture features of the device surface. The effect is not ideal in an environment where the surface texture of the cabinet equipment is single.
[0048] In this solution, regions of interest are identified based on point cloud data collected by time-of-flight sensors. Time-of-flight sensors offer the following advantages in the controlled environment of communications cabinets: They are adaptable to darkroom environments, which are typically enclosed. Their active infrared luminescence allows them to operate in complete darkness, independent of ambient light and requiring no additional lighting. They also directly acquire depth information. Time-of-flight sensors measure depth directly, avoiding complex algorithms like stereo matching. They can accurately capture geometric information even when the device surface has a single texture. Compared to lidar, time-of-flight sensors are less expensive and suitable for large-scale deployment. Furthermore, they can be conveniently deployed on the interior doorframe of a communications cabinet, enabling scanning coverage of the entire cabinet through a mobile truss.
[0049] Figure 4 A flow chart of another method for determining an area of interest for railway communication equipment provided in an embodiment of the present disclosure is shown as follows: Figure 4 As shown, in some embodiments of the present disclosure, determining multiple measurement area pairs based on multiple candidate point cloud data includes:
[0050] Step 401 : performing surface extraction processing on a plurality of candidate point cloud data to obtain surface point cloud clusters.
[0051] The surface point cloud cluster may be a set of points in the candidate point cloud data corresponding to the surface of the communication device.
[0052] In this embodiment, the region of interest determination device of the railway communication equipment can process the candidate point cloud data through a surface extraction algorithm to extract surface point cloud clusters in the candidate point cloud data. This embodiment does not limit the surface extraction algorithm. For example, multi-scale random sample consensus (RANSAC) can be used, which has high surface detection accuracy.
[0053] In some embodiments of the present disclosure, surface extraction processing is performed on a plurality of candidate point cloud data to obtain a surface point cloud cluster, including:
[0054] Step a1: for each candidate point cloud data, extract the point cloud data within a preset depth range of the candidate point cloud data to obtain filtered point cloud data.
[0055] The preset depth interval can be a pre-set interval representing the distance range between the communication cabinet surface and the time-of-flight sensor. This preset depth interval can be dynamically adjusted based on the depth layout characteristics of the communication cabinet, thereby improving the adaptability of the railway communication equipment region of interest determination method to different communication cabinets. For example, the preset depth interval can be 25 cm to 35 cm.
[0056] In this embodiment, for each candidate point cloud data, the region of interest determination device of the railway communication equipment can extract point cloud data with a depth distance within a preset depth interval to obtain filtered point cloud data.
[0057] Step a2: When the point cloud density of the filtered point cloud data is greater than a preset density threshold, point cloud clustering is performed on the filtered point cloud data to obtain candidate point cloud clusters.
[0058] The point cloud density may represent the distribution density of points in the point cloud. This embodiment does not limit the calculation method of the point cloud density. For example, the point cloud density may be the ratio of the number of points in the filtered point cloud data to the number of points in the corresponding candidate point cloud data. A candidate point cloud cluster may be a point cloud cluster whose correspondence with the device plane is to be determined.
[0059] In this embodiment, the region of interest determination device of the railway communication equipment can calculate the point cloud density of the filtered point cloud data and determine whether the point cloud density is greater than a preset density threshold. If so, the dominant plane is determined based on the filtered point cloud data, and the inner points of the dominant plane are determined as candidate point cloud clusters.
[0060] For example, the candidate point cloud data is the original depth map D(x,y). The minimum endpoint d_min in the preset depth range is 25cm, and the maximum endpoint d_max is 35cm. For each pixel in the original depth map, if d_min ≤ d(x,y) ≤ d_max, the pixel is added to the filtered point cloud data to obtain the filtered point cloud data P_surface_candidate. The point cloud density density is calculated. For example, if the number of points in the candidate point cloud data is total_pixels, the number of candidate point clouds can be calculated as: density = |P_surface_candidate| / total_pixels. If the point cloud density exceeds the preset density threshold, the current measurement area is marked as "surface occupied" and the filtered point cloud data is clustered to obtain candidate point cloud clusters.
[0061] Step a3: If the candidate point cloud cluster meets the preset surface condition, the candidate point cloud cluster is determined as a surface point cloud cluster.
[0062] The preset surface condition may be a pre-set condition for indicating that the point cloud cluster meets the plane feature. This embodiment does not limit the preset surface condition. For example, the preset surface condition may include a normal vector condition and a flatness condition. The flatness condition is also called a planarity condition.
[0063] In this embodiment, the region of interest determination device of the railway communication equipment can determine the normal vector of the candidate point cloud cluster and determine whether the normal vector is within a preset normal vector interval. The preset normal vector interval can be a preset angle interval representing the normal vector facing the time-of-flight sensor. This embodiment does not impose any restrictions on this preset normal vector interval. The region of interest determination device of the railway communication equipment can also determine the flatness of the candidate point cloud cluster and determine whether the flatness is within a preset flatness interval. The preset flatness interval can be a preset interval representing the flatness dimension meeting the plane feature. This embodiment does not impose any restrictions on this preset flatness interval. If the normal vector is within the preset normal vector interval and the flatness is within the preset flatness interval, it indicates that the candidate point cloud cluster meets the preset plane condition and the candidate point cloud cluster is determined to be a surface point cloud cluster. If the normal vector is not within the preset normal vector interval and / or the flatness is not within the preset flatness interval, it indicates that the candidate point cloud cluster is an invalid point cloud cluster.
[0064] This approach implements narrowband depth filtering of candidate point cloud data based on preset depth intervals, effectively eliminating background interference and foreground noise, enabling focused perception of the communication device surface and improving detection accuracy in areas of interest. Furthermore, point cloud density and preset surface conditions are used to further identify the point cloud clusters corresponding to the device surface, providing an accurate data foundation for subsequent point cloud cluster connectivity processing.
[0065] In step 402 , two surface point cloud clusters corresponding to adjacent measurement areas and the same communication device are divided into the same point cloud cluster pair; wherein a point cloud cluster pair includes two surface point cloud clusters.
[0066] The point cloud cluster pair can be used to record two surface point cloud clusters corresponding to the same communication device in two adjacent measurement areas.
[0067] In this embodiment, the region of interest determination device of the railway communication equipment can judge the surface point cloud clusters in adjacent measurement areas. If the surface point cloud clusters correspond to the same communication equipment, the two surface point cloud clusters are divided into the same point cloud cluster pair.
[0068] In some embodiments of the present disclosure, two surface point cloud clusters corresponding to adjacent measurement areas and the same communication device are divided into the same point cloud cluster pair, including:
[0069] In step b1, surface point cloud clusters are respectively determined as point cloud clusters to be processed, and candidate point cloud data containing the point cloud clusters to be processed are determined as point cloud data to be processed.
[0070] The point cloud cluster to be processed may be a surface point cloud cluster currently being processed, and the point cloud cluster to be processed may be any surface point cloud cluster. The point cloud data to be processed may be candidate point cloud data where the point cloud cluster to be processed is located.
[0071] In this embodiment, in order to determine the associated direction corresponding to each surface point cloud cluster, the area of interest determination device of the railway communication equipment can determine the currently processed surface point cloud cluster as the point cloud cluster to be processed, and determine the candidate point cloud data where the point cloud cluster to be processed is located as the point cloud data to be processed.
[0072] Step b2: determining a target boundary among multiple point cloud boundaries of the point cloud data to be processed according to the boundary distance between the point cloud cluster to be processed and the point cloud boundary, and determining the boundary distance between the point cloud cluster to be processed and the target boundary as the target distance.
[0073] The point cloud boundary may be a rectangular boundary corresponding to the point cloud data to be processed, and this embodiment does not impose any restrictions on the specific setting of this point cloud boundary. For example, the point cloud boundary may be the boundary of the bounding box of the point cloud data to be processed, or it may be the boundary of the measurement area where the point cloud data to be processed is located. The target boundary may be the boundary of the area spanned by the entire point cloud cluster of the communication device corresponding to the point cloud cluster to be processed. The number of target boundaries may be one or more, and this embodiment does not impose any restrictions. The target distance may be the boundary distance between the point cloud cluster to be processed and the target boundary.
[0074] In this embodiment, the region of interest determination device of the railway communication equipment can determine four point cloud boundaries of the point cloud data to be processed and calculate the boundary distance between the point cloud cluster to be processed and these four point cloud boundaries. This embodiment does not limit the method for determining these boundary distances. Furthermore, the point cloud boundary with a boundary distance within a preset distance interval is determined as the target boundary, or the point cloud boundary corresponding to the minimum boundary distance is determined as the target boundary. Furthermore, the boundary distance between this target boundary and the point cloud cluster to be processed is determined as the target distance.
[0075] Step b3: extracting boundary points in the point cloud cluster to be processed based on the distance between each point in the point cloud cluster to be processed and the target boundary.
[0076] The boundary points may be points in the point cloud cluster to be processed that are close to the target boundary. This embodiment does not limit the method for determining the boundary points.
[0077] In this embodiment, the region of interest determination device of the railway communication equipment can calculate the distance between each point in the point cloud cluster to be processed and the target boundary, and determine the point whose distance is less than a preset distance threshold as a boundary point.
[0078] Step b4: determining the association direction of the point cloud cluster to be processed according to the target distance and the boundary point, and obtaining the association direction corresponding to each surface point cloud cluster.
[0079] The associated direction may be a potential extension direction of the point cloud cluster to be processed, and the associated direction may include a combination of one or more of up, down, left, and right.
[0080] In this embodiment, the region of interest determination device of the railway communication equipment can calculate the proximity value between the to-be-processed point cloud cluster and the target boundary based on the target distance and boundary points. If the proximity value is greater than a preset proximity threshold, the orientation of the target boundary within the bounding box is determined as the associated direction corresponding to the surface point cloud cluster. For example, if the target boundary is the right boundary within the bounding box, the associated direction can be right. The proximity value can represent the degree of proximity between the point cloud cluster and the line, and this embodiment does not limit the calculation method of this proximity value.
[0081] For example, to determine the proximity of each surface point cloud cluster to the upper, lower, left, and right boundaries of the current field of view, the railway communication device's region of interest determination device may calculate the boundary distance between the center of gravity of the point cloud cluster to be processed and the field of view boundary, and determine the target boundary based on this boundary distance. Alternatively, the railway communication device's region of interest determination device may calculate the boundary distance between the bounding box boundary of the point cloud cluster to be processed and the field of view boundary, and determine the target boundary based on this boundary distance.
[0082] After determining the target boundary, the region of interest determination device of the railway communication equipment can count the percentage of boundary points in the pending point cloud cluster that are close to the target boundary. Based on this percentage and the target distance, a proximity value between the pending point cloud cluster and the target boundary is determined. This proximity value indicates whether the corresponding pending point cloud cluster is likely to cross into an adjacent measurement area. If the proximity value is greater than a proximity threshold, the orientation of the target boundary within the bounding box is determined as the associated direction corresponding to the surface point cloud cluster. Thus, the associated direction corresponding to each surface point cloud cluster is obtained.
[0083] Specifically, taking the point cloud cluster to be processed P_cluster and the point cloud boundary of the point cloud data to be processed as bounds as an example, the minimum x-direction value in the field of view boundary coordinates is bounds.x_min, the maximum x-direction value is bounds.x_max, the minimum y-direction value is bounds.y_min, and the maximum y-direction value is bounds.y_max.
[0084] First, determine the bounding box boundary bbox of the point cloud cluster to be processed. The minimum x-direction value in the bounding box boundary is bbox.x_min, the maximum x-direction value is bbox.x_max, the minimum y-direction value is bbox.y_min, and the maximum y-direction value is bbox.y_max. Further, calculate the boundary distance between the point cloud cluster to be processed and the point cloud boundary of the point cloud data to be processed in each direction. The boundary distance may include the left boundary distance dist_left, the right boundary distance dist_right, the upper boundary distance dist_top, and the lower boundary distance dist_bottom. Specifically, dist_left = bbox.x_min - bounds.x_min; dist_right = bounds.x_max - bbox.x_max; dist_top = bbox.y_min - bounds.y_min; dist_bottom = bounds.y_max - bbox.y_max.
[0085] For each direction, if the boundary distance is less than a preset distance threshold, a distance component value is determined based on the boundary distance and the preset distance threshold. The distance component value can be the difference between 1 and the quotient between the boundary distance and the preset distance threshold. If the difference is within a difference interval (e.g., a numerical interval close to 1), the orientation of the target boundary within the bounding box is determined as the associated direction corresponding to the surface point cloud cluster.
[0086] Step b5: determine the adjacent directions of two adjacent measurement areas. If there are two surface point cloud clusters whose associated directions match the adjacent directions in the two adjacent measurement areas, the two surface point cloud clusters are divided into the same point cloud cluster pair.
[0087] The adjacent direction may represent the positional relationship between two adjacent measurement areas, and the adjacent direction may include vertically adjacent and horizontally adjacent.
[0088] In this embodiment, the association direction of each surface point cloud cluster is determined by determining the surface point cloud clusters as the point cloud clusters to be processed. Furthermore, the region of interest determination device of the railway communication equipment can traverse the adjacent measurement areas in pairs in multiple measurement areas. If the adjacent directions of the measurement areas correspond to the association directions of the surface point cloud clusters in the two measurement areas, the surface point cloud clusters are divided into the same point cloud cluster pair. For example, if in two measurement areas adjacent to each other, the upper measurement area has a surface point cloud cluster with an association direction including the lower one, and the lower measurement area has a surface point cloud cluster with an association direction including the upper one, then the two surface point cloud clusters are divided into the same point cloud cluster pair. If in two measurement areas adjacent to each other, the left measurement area has a surface point cloud cluster with an association direction including the right one, and the right measurement area has a surface point cloud cluster with an association direction including the left one, then the two surface point cloud clusters are divided into the same point cloud cluster pair.
[0089] Step 403: Determine a measurement region pair based on the point cloud cluster pair.
[0090] In this embodiment, two measurement regions where two surface point cloud clusters in a point cloud cluster pair are located are determined, and a measurement region pair including the two measurement regions is determined.
[0091] In the above scheme, by evaluating the boundary proximity of point cloud clusters and analyzing the proximity of point cloud clusters to the boundaries, the potential correlation with point cloud clusters in adjacent measurement areas is judged. The correlation relationship between point cloud clusters across measurement areas is accurately predicted, and it is determined whether point cloud clusters distributed in different measurement areas belong to the same communication device surface, providing an accurate basis for subsequent confirmation scanning.
[0092] Step 102: Determine a plurality of connected boundaries corresponding to a plurality of measurement area pairs; wherein a connected boundary is a region boundary between two measurement areas included in a measurement area pair.
[0093] In an embodiment of the present disclosure, a region of interest determination device of a railway communication device can, for each measurement area pair, determine the two measurement areas included in the measurement area pair and determine the region boundary between the two measurement areas as a connected boundary. For example, if multiple measurement areas are distributed in a grid pattern, the region boundary shared by the two measurement areas in the measurement area pair can be used as the connected boundary.
[0094] Step 103 : acquiring a plurality of boundary point cloud data sets generated by scanning and measuring a plurality of connected boundaries, and determining the connected boundaries corresponding to the boundary point cloud data sets that satisfy the surface continuity condition as target boundaries.
[0095] Among them, the scanning measurement can be the measurement of point cloud data of a connected boundary in the direction of the connected boundary with a preset scanning interval. A boundary point cloud data group can be a collection of multiple boundary point cloud data obtained by scanning and measuring a connected boundary. This embodiment does not limit the number of boundary point cloud data included in a boundary point cloud data group. For example, it can be 5. The boundary point cloud data can be point cloud data obtained by a single measurement of the connected boundary. It can be understood that the boundary point cloud data is point cloud data across the measurement area. The surface continuity condition can be a pre-set condition that characterizes the continuity of the surface in the multi-frame point cloud data obtained by continuous measurement. The surface continuity condition can be set according to user needs, etc., and is not limited in this embodiment. The target boundary can be the regional boundary connecting the point cloud clusters in the two measurement areas.
[0096] In an embodiment of the present disclosure, for each connected boundary, the region of interest determination device of the railway communication equipment can send the identifier of the connected boundary to the control unit corresponding to the mobile truss. The control unit performs path planning of the time-of-flight sensor based on the identifier of the connected boundary to obtain a planned path. For example, for a horizontal connected boundary, a linear scanning measurement can be performed left and right along the connected boundary. For a vertical connected boundary, a linear scanning measurement can be performed up and down along the connected boundary. The mobile truss carries a time-of-flight sensor to perform scanning measurements according to the planned path to obtain a boundary point cloud data group corresponding to each connected boundary. Furthermore, the region of interest determination device of the railway communication equipment can determine whether the boundary point cloud data meets the surface continuity condition. If so, the connected boundary is determined as the target boundary.
[0097] In some embodiments of the present disclosure, determining a connected boundary corresponding to a boundary point cloud data set that satisfies a surface continuity condition as a target boundary includes:
[0098] Surface extraction processing is performed on multiple boundary point cloud data in the boundary point cloud data group to obtain multiple boundary point cloud clusters; if the depth information of the multiple boundary point cloud clusters meets the depth continuity condition, and the normal vector information of the multiple boundary point cloud clusters meets the normal vector continuity condition, then the connected boundary corresponding to the boundary point cloud data group is determined as the target boundary.
[0099] The boundary point cloud cluster can be a point cloud cluster corresponding to a surface extracted from the boundary point cloud data. Depth information can be used to record the overall depth of points in the boundary point cloud cluster. The depth continuity condition can be expressed in the depth dimension, where multiple frame boundary point cloud clusters are acquired based on the same surface. Normal vector information can be used to record the normal vector of the plane fitted to the boundary point cloud cluster. The normal vector continuity condition can be expressed in the normal vector dimension, where multiple frame boundary point cloud clusters are acquired based on the same surface.
[0100] In this embodiment, for each boundary point cloud data group, the region of interest determination device of the railway communication equipment can perform surface extraction processing on the multiple boundary point cloud data in the boundary point cloud data group to obtain multiple boundary point cloud clusters. It can be understood that since the boundary point cloud data in a boundary point cloud data group is obtained by scanning measurement, the boundary point cloud data has a sequential relationship, and therefore the boundary point cloud clusters determined based on the boundary point cloud data also have a sequential relationship. Furthermore, the region of interest determination device of the railway communication equipment can determine the depth information and normal vector information of the boundary point cloud cluster. If the difference between the depth information of two adjacent boundary point cloud clusters in the sequential relationship is less than a preset depth difference threshold, it is determined that the depth information of the multiple boundary point cloud clusters meets the depth continuity condition. This embodiment does not limit the preset depth difference threshold. For example, the preset depth difference threshold can be 3mm.
[0101] If the difference between the normal vector information of two adjacent boundary point cloud clusters in a sequential relationship is less than a preset normal vector difference threshold, the normal vector information of the multiple boundary point cloud clusters is determined to meet the normal vector continuity condition. This embodiment does not impose any restrictions on this normal vector difference threshold; for example, the normal difference threshold can be 5°. If both the depth continuity condition and the normal vector continuity condition are met, it indicates that the point cloud data across the measurement area has planar continuity, and the point cloud clusters in the point cloud data belong to the same communication device surface. The connected boundary corresponding to the boundary point cloud data group is then determined as the target boundary.
[0102] Optionally, after determining the target boundary, the geometric transformation relationship between the surface point cloud clusters connected by the target boundary can be calculated, the global consistency constraint can be optimized, and the transformation parameters based on the splicing transformation of the surface point cloud clusters can be output.
[0103] In the above scheme, a special scanning path is designed to perform scanning measurement on the connected boundaries across the measurement area, and the plane continuity detection is performed on the scanned point cloud data. This can accurately identify the plane continuity changes across the measurement area, realize the determination of the affiliation of point cloud clusters in different measurement areas, and improve the accuracy and reliability of the point cloud cluster association confirmation.
[0104] Step 104 : grouping the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups.
[0105] A surface point cloud cluster group may be a set of surface point cloud clusters corresponding to a surface of a communication device.
[0106] In the embodiment of the present disclosure, the region of interest determination device of the railway communication equipment can determine the surface point cloud clusters connected by the target boundary, and divide the surface point cloud clusters connected as a whole into corresponding surface point cloud cluster groups.
[0107] In some embodiments of the present disclosure, surface point cloud clusters extracted from multiple candidate point cloud data are grouped according to target boundaries to obtain multiple surface point cloud cluster groups, including: assigning the same device identifier to surface point cloud clusters connected by target boundaries; and dividing surface point cloud clusters corresponding to the same device identifier into the same surface point cloud cluster group.
[0108] The device identifier can be used to uniquely identify the communication device.
[0109] In this embodiment, the region of interest determination device of the railway communication equipment can assign the same device identifier to two surface point cloud clusters corresponding to the target boundary, and group the surface point cloud clusters corresponding to the same device identifier into the same surface point cloud cluster group, thereby achieving chain grouping of surface point cloud clusters based on the target boundary. For example, if one target boundary corresponds to surface point cloud cluster A and surface point cloud cluster B, and another target boundary corresponds to surface point cloud cluster B and surface point cloud cluster C, then the surface point cloud cluster group includes surface point cloud cluster A, surface point cloud cluster B, and surface point cloud cluster C.
[0110] In the above solution, the association relationship between surface point cloud clusters across measurement areas based on target boundary representation is realized, and it is determined that surface point cloud clusters distributed in multiple measurement areas belong to the same communication device.
[0111] Step 105 : determining the region of interest of the communication device according to the surface point cloud cluster group.
[0112] The region of interest may be a region enclosed by a communication device. This embodiment does not limit the shape of the region of interest. For example, the region of interest may be rectangular.
[0113] In this embodiment, the ROI determination device for railway communication equipment can determine a frame of a preset shape enclosing multiple surface point cloud clusters based on their positions within each surface point cloud cluster, and then define the area within the frame as the ROI. Thus, based on the surface point cloud clusters identified across measurement areas for the same communication device, the accurate ROI corresponding to the complete communication device is determined. Furthermore, the splicing of surface point cloud clusters across measurement areas ensures geometric accuracy and consistent ROI calibration, meeting the accuracy requirements of subsequent visual inspection.
[0114] The method for determining the region of interest of railway communication equipment provided by the embodiment of the present disclosure includes: obtaining multiple candidate point cloud data generated by performing regional measurement on a communication cabinet, and determining multiple measurement area pairs based on the multiple candidate point cloud data; wherein, multiple communication devices are arranged in the communication cabinet, the candidate point cloud data correspond to the measurement areas one by one, and the measurement area pair includes two adjacent measurement areas where the same communication device exists; determining multiple connected boundaries corresponding to the multiple measurement area pairs; wherein the connected boundary is the region boundary between the two measurement areas included in the measurement area pair; obtaining multiple boundary point cloud data groups generated by scanning measurement of the multiple connected boundaries, and determining the connected boundaries corresponding to the boundary point cloud data groups that meet the surface continuity condition as target boundaries; grouping and processing the surface point cloud clusters extracted from the multiple candidate point cloud data according to the target boundaries to obtain multiple surface point cloud cluster groups; and determining the region of interest of the communication equipment based on the surface point cloud cluster groups.
[0115] By adopting the above technical solution, candidate point cloud data is obtained by grid scanning the communication cabinet, and adjacent pairwise measurement areas containing the same communication equipment are determined. The connected boundary between the two paired measurement areas is scanned and measured to obtain multiple boundary point cloud data. When multiple boundary point cloud data groups meet the plane continuity condition, the connected boundary is determined as the target boundary, and point cloud clusters are aggregated based on the target boundary and the area of interest of the communication equipment is determined. By scanning the connected boundary that is potentially the splicing boundary of the point cloud cluster and determining that the scanned boundary point cloud data meets the plane continuity condition, it is indicated that the connected boundary is actually the boundary of the area crossed by the communication equipment. Then, the point cloud clusters across the field of view are aggregated and the area of interest is identified based on the connected boundary, thereby realizing the accurate and efficient determination of the area of interest of the communication equipment when the same communication equipment appears in multiple adjacent fields of view.
[0116] In addition, the cross-field of view problem caused by the mismatch between the field of view size of the time-of-flight sensor and the size of the communication equipment is solved, and the high-precision, high-efficiency and low-cost automatic generation of the region of interest is achieved. The problem of remote monitoring and deployment of communication equipment is solved, and the foundation is created for the automated inspection of communication equipment.
[0117] Optionally, in some embodiments of the present disclosure, after determining the region of interest, the size of the region of interest can also be verified. Specifically, the region of interest determination device of the railway communication equipment can determine whether the width of the region of interest is within the preset width interval, and determine whether the height of the region of interest is within the height interval. If both are true, it means that the region of interest has passed the size verification. This embodiment does not limit the preset width interval and the preset height interval, and the preset width interval and the preset height interval can be set according to the standard cabinet rules. For example, the preset width interval can be 38cm-45cm, and the preset height interval can be 4.5cm-90cm. Thereby, the size of the region of interest meets the specifications of the standard cabinet equipment.
[0118] If the size of the region of interest is determined to be abnormal, the boundaries of the region of interest can be adjusted. The region of interest determination device of the railway communication device can determine whether the regions of interest of adjacent devices overlap. If so, the overlapping regions of interest are adjusted to obtain non-overlapping regions of interest. The corrected regions of interest are output as a region of interest list.
[0119] Next, the method for determining the region of interest in the embodiment of the present disclosure is further described through a specific example.
[0120] First, the system is initialized. Specifically, the internal parameters of the time-of-flight sensor and the coordinate system of the mobile truss are calibrated. The grid scanning spacing, preset depth intervals, and various thresholds are set to confirm the range of motion of the mobile truss and the installation position of the time-of-flight sensor. Furthermore, the scan is executed. Specifically, the control unit sends a movement instruction to the execution unit, which controls the mobile truss to move to the specified measurement position. The time-of-flight sensor collects candidate point cloud data, and the control unit performs filtering and surface occupancy determination based on the preset depth interval. This process is repeated until the entire communication cabinet is scanned. Connectivity analysis is performed on the measurement areas marked as surface occupied. Plane detection is performed on each surface occupied measurement area to determine planar point cloud clusters. Connected boundaries are determined based on the planar point cloud clusters. A fine scanning path is planned for the connected boundaries. Guided linear scanning is performed based on this scanning path. Depth continuity and normal vector continuity detection are performed in real time. The connected boundaries that pass these two detections are determined as target boundaries, and the surface point cloud clusters connected to the target boundaries are optimized.
[0121] Furthermore, a corresponding region of interest is generated for each communication device, and the area and execution degree of the region of interest are determined, and a plan view of the communication cabinet is generated based on the region of interest. Figure 5 A schematic diagram of a communication cabinet provided in an embodiment of the present disclosure is shown in FIG. Figure 5As shown, the one-sided diagram depicts rectangular boxes representing each communication device based on the location of the region of interest of the communication device. Each rectangular box is labeled with the name of the communication device. The communication device names may include: baseband processing unit-01, baseband processing unit-02, remote radio unit-01, remote radio unit-02, transmission unit-01, transmission unit-02, power supply unit-01, power supply unit-02, patch panel-01, patch panel-02, and environmental monitoring unit. Optionally, the Internet Protocol (IP) address of the communication device may also be labeled within each rectangular box.
[0122] The method for determining the region of interest of the above-mentioned railway communication equipment provided in the embodiment of the present disclosure replaces the manual region of interest calibration method in the internal environment of the communication cabinet by using a time-of-flight sensor on a mobile truss deployed on a door frame, thereby realizing the splicing of surface point cloud clusters corresponding to the device surface across the field of view and high-precision automatic calibration of the region of interest.
[0123] The ROI calibration time for a single communications cabinet has been reduced from 2-4 hours manually to 30-45 minutes automatically, improving efficiency. This reduces the need for on-site technicians and lowers maintenance costs, making it particularly suitable for large-scale or dispersed deployments such as 5G base stations, railway communications, power communications, and remote areas. Remote real-time monitoring provides a precise basis for regions of interest, improving the timeliness of equipment status changes and fault detection. Furthermore, the mobile truss is installed in the door frame of the communications cabinet, making it easy to deploy, allowing for quick assembly and disassembly, and mobile use, with low maintenance costs. Leveraging the active illumination of the time-of-flight sensor, it operates stably in the darkroom environment of the communications cabinet, unaffected by changes in external lighting.
[0124] By analyzing the proximity of point cloud clusters to boundaries, the potential association with point cloud clusters in adjacent measurement areas is determined, providing a basis for stitching point cloud clusters across measurement areas. Point cloud clusters distributed in adjacent measurement areas are accurately identified as belonging to the same communication device surface, establishing cross-measurement area association relationships. This avoids misassignment of surface point cloud clusters and enables cross-field visual identification of surface point cloud clusters. A specialized scanning path and detection algorithm are designed for connected boundaries across measurement areas to confirm the affiliation of point cloud clusters in different measurement areas. Point cloud clusters across multiple measurement areas are correctly stitched together to form a complete device surface, generating accurate region-of-interest boundaries. To address the limitation of the time-of-flight sensor's field of view, which cannot fully cover the communication device, efficient and accurate ROI generation is achieved through optimized scanning strategies and stitching algorithms. This ensures that the geometric accuracy of the stitching process for surface point cloud clusters across measurement areas and the consistency of region-of-interest calibration between different devices meet the accuracy requirements of subsequent visual inspection. This improves the stability of the subsequent inspection algorithm performance.
[0125] Figure 6 This is a schematic diagram of the structure of a device for determining an area of interest of a railway communication device provided by an embodiment of the present disclosure. The device can be implemented by software and / or hardware, or integrated into an electronic device. Figure 6 As shown, the device for determining the region of interest of the railway communication equipment includes:
[0126] The region pair determination module 601 is configured to obtain a plurality of candidate point cloud data generated by performing region-by-region measurement on a communication cabinet, and determine a plurality of measurement region pairs based on the plurality of candidate point cloud data; wherein the communication cabinet contains a plurality of communication devices, the candidate point cloud data corresponds one-to-one to the measurement regions, and the measurement region pairs include two adjacent measurement regions containing the same communication device;
[0127] A first boundary determination module 602 is configured to determine a plurality of connected boundaries corresponding to the plurality of measurement area pairs; wherein the connected boundary is a region boundary between two measurement areas included in a measurement area pair;
[0128] The second boundary determination module 603 is configured to obtain a plurality of boundary point cloud data groups generated by scanning and measuring the plurality of connected boundaries, and determine the connected boundaries corresponding to the boundary point cloud data groups that meet the surface continuity condition as target boundaries;
[0129] A grouping module 604 is configured to group the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups;
[0130] The region determination module 605 is configured to determine a region of interest of the communication device according to the surface point cloud cluster group.
[0131] Optionally, the candidate point cloud data is collected by a time-of-flight sensor, which is fixed on a mobile truss. The movement plane of the mobile truss is arranged opposite to the operating surface of the communication cabinet. The mobile truss is used to move the time-of-flight sensor to a preset measurement position, and the measurement position is arranged in a two-dimensional array.
[0132] Optionally, determining a plurality of measurement area pairs based on the plurality of candidate point cloud data includes:
[0133] Performing surface extraction processing on the plurality of candidate point cloud data to obtain surface point cloud clusters;
[0134] Dividing two surface point cloud clusters whose corresponding measurement areas are adjacent and correspond to the same communication device into the same point cloud cluster pair; wherein one point cloud cluster pair includes two surface point cloud clusters;
[0135] The measurement region pair is determined according to the point cloud cluster pair.
[0136] Optionally, performing surface extraction processing on the plurality of candidate point cloud data to obtain a surface point cloud cluster includes:
[0137] For each candidate point cloud data, extracting point cloud data within a preset depth interval of the candidate point cloud data to obtain filtered point cloud data;
[0138] When the point cloud density of the filtered point cloud data is greater than a preset density threshold, performing point cloud clustering on the filtered point cloud data to obtain candidate point cloud clusters;
[0139] If the candidate point cloud cluster meets the preset surface condition, the candidate point cloud cluster is determined as the surface point cloud cluster.
[0140] Optionally, dividing two surface point cloud clusters corresponding to adjacent measurement areas and corresponding to the same communication device into the same point cloud cluster pair includes:
[0141] Determining the surface point cloud clusters as to-be-processed point cloud clusters respectively, and determining the candidate point cloud data containing the to-be-processed point cloud clusters as to-be-processed point cloud data;
[0142] Determining a target boundary among the plurality of point cloud boundaries of the point cloud data to be processed according to a boundary distance between the point cloud cluster to be processed and the point cloud boundary, and determining the boundary distance between the point cloud cluster to be processed and the target boundary as a target distance;
[0143] Extracting boundary points in the point cloud cluster to be processed according to the distance between each point in the point cloud cluster to be processed and the target boundary;
[0144] Determining the association direction of the point cloud cluster to be processed according to the target distance and the boundary point, and obtaining the association direction corresponding to each of the surface point cloud clusters;
[0145] The adjacent directions of two adjacent measurement areas are determined, and if there are two surface point cloud clusters whose associated directions match the adjacent directions in the two adjacent measurement areas, the two surface point cloud clusters are divided into the same point cloud cluster pair.
[0146] Optionally, determining the connected boundary corresponding to the boundary point cloud data set that satisfies the surface continuity condition as the target boundary includes:
[0147] performing surface extraction processing on a plurality of boundary point cloud data in the boundary point cloud data group to obtain a plurality of boundary point cloud clusters;
[0148] If the depth information of the plurality of boundary point cloud clusters satisfies a depth continuity condition, and the normal vector information of the plurality of boundary point cloud clusters satisfies a normal vector continuity condition, the connected boundary corresponding to the boundary point cloud data group is determined as the target boundary.
[0149] Optionally, grouping the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups includes:
[0150] Assigning the same device identifier to surface point cloud clusters connected by the target boundary;
[0151] The surface point cloud clusters corresponding to the same device identification are divided into the same surface point cloud cluster group.
[0152] It should be noted that Figure 6 The shown device for determining the region of interest of the railway communication equipment can execute the various steps in the embodiment of the method for determining the region of interest of the railway communication equipment, and realize the various processes and effects in the embodiment of the method for determining the region of interest of the railway communication equipment, which will not be described in detail here.
[0153] Figure 7 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. Figure 7 As shown, electronic device 700 includes one or more processors 701 and memory 702 .
[0154] The processor 701 may be a central processing unit (CPU) or other forms of processing units having the capability of determining an area of interest and / or executing instructions for railway communication equipment, and may control other components in the electronic device 700 to perform desired functions.
[0155] The memory 702 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 701 may execute the program instructions to implement the region of interest determination method for railway communication equipment according to the embodiments of the present disclosure described above and / or other desired functions. The computer-readable storage medium may also store various contents, such as input signals, signal components, and noise components.
[0156] In one example, the electronic device 700 may further include an input device 703 and an output device 704 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).
[0157] In addition, the input device 703 may also include, for example, a keyboard, a mouse, and the like.
[0158] The output device 704 can output various information to the outside, including determined distance information, direction information, etc. The output device 704 can include, for example, a display, a speaker, a printer, a communication network and its connected remote output device, etc.
[0159] Of course, to simplify, Figure 7 Only some of the components related to the present disclosure in the electronic device 700 are shown, and components such as buses, input / output interfaces, etc. are omitted. In addition, the electronic device 700 may further include any other appropriate components according to specific application scenarios.
[0160] In addition to the above methods and devices, the embodiments of the present disclosure may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the method for determining the area of interest of railway communication equipment provided by the embodiments of the present disclosure.
[0161] The computer program product may be written in any combination of one or more programming languages to implement the operations of the disclosed embodiments, including object-oriented programming languages such as Java, C++, and conventional procedural programming languages such as C or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.
[0162] In addition, the embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, enables the processor to execute the method for determining the region of interest of railway communication equipment provided by the embodiment of the present disclosure.
[0163] The computer-readable storage medium may be any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may include, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.
[0164] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.
[0165] The foregoing description is intended only to provide specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments described herein, but rather to be construed in the broadest manner consistent with the principles and novel features disclosed herein.
Claims
1. A method for determining an area of interest for railway communication equipment, characterized in that: include: Acquire multiple candidate point cloud data generated by performing regional measurement on the communication cabinet, and determine multiple measurement area pairs based on the multiple candidate point cloud data; wherein the communication cabinet has multiple communication devices arranged therein, the candidate point cloud data corresponds one-to-one to the measurement areas, and the measurement area pairs include two adjacent measurement areas in which the same communication device is located; Determining a plurality of connected boundaries corresponding to the plurality of measurement area pairs; wherein the connected boundary is a region boundary between two measurement areas included in the measurement area pair; Acquire a plurality of boundary point cloud data groups generated by scanning and measuring the plurality of connected boundaries, and determine the connected boundaries corresponding to the boundary point cloud data groups that meet the surface continuity condition as target boundaries; Grouping the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups; A region of interest of the communication device is determined based on the surface point cloud cluster group.
2. The method according to claim 1, characterized in that The candidate point cloud data is collected by a time-of-flight sensor, which is fixed on a mobile truss. The movement plane of the mobile truss is arranged opposite to the operating surface of the communication cabinet. The mobile truss is used to move the time-of-flight sensor to a preset measurement position, and the measurement position is arranged in a two-dimensional array.
3. The method according to claim 1, characterized in that The step of determining a plurality of measurement area pairs based on the plurality of candidate point cloud data comprises: Performing surface extraction processing on the plurality of candidate point cloud data to obtain surface point cloud clusters; Dividing two surface point cloud clusters whose corresponding measurement areas are adjacent and correspond to the same communication device into the same point cloud cluster pair; wherein one point cloud cluster pair includes two surface point cloud clusters; The measurement region pair is determined according to the point cloud cluster pair.
4. The method according to claim 3, characterized in that The performing surface extraction processing on the plurality of candidate point cloud data to obtain a surface point cloud cluster includes: For each candidate point cloud data, extracting point cloud data within a preset depth interval of the candidate point cloud data to obtain filtered point cloud data; When the point cloud density of the filtered point cloud data is greater than a preset density threshold, performing point cloud clustering on the filtered point cloud data to obtain candidate point cloud clusters; If the candidate point cloud cluster meets the preset surface condition, the candidate point cloud cluster is determined as the surface point cloud cluster.
5. The method according to claim 3, characterized in that The step of dividing two surface point cloud clusters corresponding to adjacent measurement areas and the same communication device into the same point cloud cluster pair includes: Determining the surface point cloud clusters as to-be-processed point cloud clusters respectively, and determining the candidate point cloud data containing the to-be-processed point cloud clusters as to-be-processed point cloud data; Determining a target boundary among the plurality of point cloud boundaries of the point cloud data to be processed according to a boundary distance between the point cloud cluster to be processed and the point cloud boundary, and determining the boundary distance between the point cloud cluster to be processed and the target boundary as a target distance; Extracting boundary points in the point cloud cluster to be processed according to the distance between each point in the point cloud cluster to be processed and the target boundary; Determining the association direction of the point cloud cluster to be processed according to the target distance and the boundary point, and obtaining the association direction corresponding to each of the surface point cloud clusters; The adjacent directions of two adjacent measurement areas are determined, and if there are two surface point cloud clusters whose associated directions match the adjacent directions in the two adjacent measurement areas, the two surface point cloud clusters are divided into the same point cloud cluster pair.
6. The method according to claim 1, characterized in that The step of determining the connected boundary corresponding to the boundary point cloud data set that satisfies the surface continuity condition as the target boundary includes: performing surface extraction processing on a plurality of boundary point cloud data in the boundary point cloud data group to obtain a plurality of boundary point cloud clusters; If the depth information of the plurality of boundary point cloud clusters satisfies a depth continuity condition, and the normal vector information of the plurality of boundary point cloud clusters satisfies a normal vector continuity condition, the connected boundary corresponding to the boundary point cloud data group is determined as the target boundary.
7. The method according to claim 1, characterized in that The grouping of the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups includes: Assigning the same device identifier to surface point cloud clusters connected by the target boundary; The surface point cloud clusters corresponding to the same device identification are divided into the same surface point cloud cluster group.
8. A device for determining a region of interest for railway communication equipment, characterized in that: include: an area pair determination module, configured to obtain a plurality of candidate point cloud data generated by performing area-by-area measurement on a communication cabinet, and determine a plurality of measurement area pairs based on the plurality of candidate point cloud data; wherein the communication cabinet contains a plurality of communication devices, the candidate point cloud data corresponds one-to-one to the measurement areas, and the measurement area pairs include two adjacent measurement areas containing the same communication device; A first boundary determination module is configured to determine a plurality of connected boundaries corresponding to the plurality of measurement area pairs; wherein the connected boundary is a region boundary between two measurement areas included in the measurement area pair; a second boundary determination module, configured to obtain a plurality of boundary point cloud data groups generated by scanning and measuring the plurality of connected boundaries, and determine the connected boundaries corresponding to the boundary point cloud data groups that meet the surface continuity condition as target boundaries; a grouping module, configured to group the surface point cloud clusters extracted from the plurality of candidate point cloud data according to the target boundary to obtain a plurality of surface point cloud cluster groups; The region determination module is configured to determine a region of interest of the communication device according to the surface point cloud cluster group.
9. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the method for determining an area of interest of a railway communication device as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that The storage medium stores a computer program, and the computer program is used to execute the method for determining an area of interest of a railway communication device according to any one of claims 1 to 7.
Citation Information
Patent Citations
Sensor data fusion method and device, electronic equipment and storage medium
CN114528941A
Patrol robot hidden area autonomous cognition algorithm based on 3D laser radar
CN119538969A