Column positioning identification method based on spatial information

Abstract

The invention relates to the technical field of column positioning identification, in particular to a column positioning identification method based on spatial information, which automatically acquires spatial information through a robot, automatically acquires building spatial edge data in combination with a moving coordinate and a distance measuring sensor, and constructs an efficient line patrol path. The robot is automatically calibrated through spatial information, accurate positioning of multiple cylinders is achieved, the measurement time is remarkably shortened, and the operation efficiency is improved. The rasterization processing and vector quantization technology is used for segmentation and coding of the barrier edge, the stability of spatial information is enhanced, the column type feature library is combined for comparison, the phenomena of misjudgment and missed judgment are effectively reduced, and the accuracy of column positioning is ensured. By constructing a full-path diagram and mapping the column shape and positioning into an indoor space model, the automation of identifying the column shape and constructing the model is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of column positioning and identification, and in particular to a column positioning and identification method based on spatial information. Background Art

[0002] In indoor buildings, especially in large rough-shell building environments, columns are key elements that support the building structure. In such environments, accurate identification and positioning of columns are important links that cannot be ignored in the construction, monitoring, and subsequent design and decoration process. The traditional column positioning method mainly relies on manual measurement. Surveyors use tools such as tape measures and laser rangefinders to locate columns by calibrating reference points, manually calculating and recording the coordinate information of the columns. Although this method can complete the task to a certain extent, it has the following obvious shortcomings:

[0003] Manual measurement requires surveyors to confirm the position of columns one by one and record and calculate data on site. For large buildings or complex indoor environments, this process is extremely time-consuming and the measurement workload is huge. Since manual operation inevitably has errors, especially in long-distance measurement and inconsistent technical levels of different operators, the measurement accuracy may be affected, resulting in deviations in the column position record. Manual measurement requires personnel to repeatedly measure and confirm, and requires post-integration and processing to generate a spatial model of the building. This process may cause data lag and reduced accuracy, and affect subsequent construction or architectural design adjustments. Summary of the invention

[0004] In order to solve the above problems, the present invention provides a column positioning and identification method based on spatial information.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] A column positioning and identification method based on spatial information comprises the following steps:

[0007] S1, making the robot automatically move around the edge of the indoor plane, and obtaining the edge range of the indoor space according to the moving coordinates and ranging sensors;

[0008] S2, constructing a patrol path inward based on the edge range of the indoor space, controlling the robot to move according to the patrol path and perform detour and obstacle avoidance through real-time distance measurement, and calibrating the actual moving path within the edge range of the indoor space to obtain a full path map; the detour and obstacle avoidance detours by maintaining a constant distance from the obstacle until returning to the patrol path;

[0009] S3. Based on the full path diagram, the obstacle area that the patrol line path cannot pass through is identified, and the edge of the obstacle area is rasterized to be divided into several sub-areas; a coding value is assigned to the sub-area according to the edge direction data of the obstacle area contained in the sub-area, and the coding value array is obtained by arranging the sub-areas according to the adjacent relationship, and the coding value array is traversed through the preset mean window to remove duplicates to obtain a feature value array;

[0010] S4, traversing the characteristic value array and the preset column characteristic value library to determine the column type;

[0011] S5. Construct an indoor space plan based on the edge range of the indoor space and the obstacle area of ​​the determined column type, map the column into the indoor space plan, and visualize the semantic information of the column.

[0012] Furthermore, the robot includes a positioning module, a distance measurement module and a movement module;

[0013] The mobile module is used to control the steering and movement of the robot and record its parameters;

[0014] The distance measuring module is used to obtain the distance between the robot body and other objects through a laser distance measuring sensor;

[0015] The positioning module is used to calculate the robot displacement data through GPS to verify the parameters of steering and movement.

[0016] Furthermore, the method of causing the robot to automatically move around the edge of an indoor plane includes the following steps:

[0017] Based on the SLAM algorithm, the edge contour information of the indoor space is obtained and the initial position of the robot is determined;

[0018] Use the A* algorithm to plan the optimal path for the robot to move along the edge of the room based on the edge contour, and adjust the path in real time to maintain a constant distance from the edge;

[0019] Based on the Hough transform algorithm, the geometric features of the indoor plane edge are extracted from the ranging data in real time, and the robot is controlled to avoid obstacles according to the ranging data;

[0020] The iterative closest point algorithm is used to correct the path after the robot completes the detour to close the path loop.

[0021] Furthermore, obtaining the edge range of the indoor space according to the mobile coordinates and the ranging sensor includes:

[0022] Use the positioning module on the robot to obtain its current position, and determine the reference coordinates of the edge area in combination with the initial position indoors;

[0023] Obtain the distance data between the robot and the wall in real time and generate a set of distance measurement points at the indoor edge;

[0024] The least square fitting algorithm is used to process the data of the ranging point set to fit the geometric shape representing the indoor edge;

[0025] Based on the fitted edge geometry, the effective edge range of the indoor space is calculated and mapped to the indoor space coordinate system.

[0026] Furthermore, constructing a patrol path inwardly based on the edge range of the indoor space includes generating a patrol path based on the edge contour of the indoor space through a path planning algorithm, and the patrol path is constructed gradually inwardly from the edge in a round-trip form and covers the entire indoor plane space.

[0027] Furthermore, S3 comprises the following steps:

[0028] Based on the edge of the obstacle area in the full path graph, edge data is extracted by a boundary detection algorithm, and the edge of the obstacle area is divided into a plurality of rectangular grids according to a predetermined grid unit size, each of which has a unique index identifier;

[0029] For each rectangular grid, according to the edge direction information of the obstacle area contained therein, vector quantization is applied to divide the edge direction data into several angle intervals and set encoding values, wherein the edge direction data includes angle size and line direction;

[0030] According to the spatial position relationship of adjacent rectangular grids, an adjacency matrix between grid cells is constructed, and the code values ​​are arranged according to the spatial distribution order of the grid cells in the adjacent matrix to form a code value array;

[0031] Traversing the encoding value array through a preset first mean window, removing encoding values ​​in the preset mean window that are equal to the mean of the preset mean window, to obtain a semi-deduplicated array; the size of the preset first mean window is three encoding value units;

[0032] The encoding value array is traversed through the preset second mean window, and the encoding values ​​in the preset mean window that are equal to the mean of the preset mean window are removed to obtain the characteristic value array, and the size of the preset first mean window is two encoding value units.

[0033] Further, the S4 comprises the following steps:

[0034] Perform equivalent sorting according to the eigenvalue array to generate several equivalent eigenvalue sequences;

[0035] The preset columnar feature value library is compared with the equivalent feature value sequence and the similarity value is calculated. If the similarity value is lower than the preset similarity threshold, no processing is performed;

[0036] If the similarity value is greater than or equal to a preset similarity threshold, the obstacle area is marked as a column type described by a preset column feature value library.

[0037] Furthermore, the formula for the equivalent sorting is as follows:

[0038]

[0039] Among them, S is the eigenvalue array, containing n eigenvalues; k is the number of rotations; i is the index; S k It is a new array obtained by rotating the eigenvalue array S k times, that is, the rotated eigenvalue sequence; mod is the modular operation.

[0040] Further, the S5 comprises the following steps:

[0041] Based on the edge range and full path map of the indoor space, the indoor space plan is constructed using GIS algorithm;

[0042] According to the obstacle area corresponding to the column, its coordinates and range in the plan view are determined and mapped in the indoor space plan view;

[0043] Different types of columns are semantically labeled using identifiers.

[0044] The beneficial effects of the present invention are as follows: the present invention realizes the collection of spatial information through a robot, automatically obtains the edge data of the building space by combining mobile coordinates with a distance sensor, and constructs an efficient line patrol path based on this information. Through the automatic calibration of spatial information, the robot can continuously and seamlessly complete the precise positioning of multiple columns, effectively shorten the measurement time and improve the overall operating efficiency of the system. Through the precise distance sensor and spatial information feedback, the spatial geometric characteristics of the column can be dynamically captured. At the same time, the rasterization processing technology is used to refine the edge of the obstacle, and the edge data is encoded by vector quantization. The encoding processing greatly enhances the stability of the spatial information, and combined with the comparison with the column feature library, it effectively reduces the misjudgment and missed judgment in the column positioning, and ensures the accuracy and reliability of the measurement. In this scheme, the spatial positioning of the column is completed by the robot, and the full path map can be constructed in real time during the cruise process, and it can be mapped to the indoor space model, so as to realize the automation of column positioning and model construction, effectively improve the accuracy of column recognition and positioning and save human resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of the steps of a column positioning and identification method based on spatial information in the present invention.

[0046] Figure 2It is a step flow chart of step S3 in the present invention. DETAILED DESCRIPTION

[0047] See also Figure 1-2 As shown, the present invention relates to a column positioning and identification method based on spatial information, comprising the following steps:

[0048] S1, making the robot automatically move around the edge of the indoor plane, and obtaining the edge range of the indoor space according to the moving coordinates and ranging sensors;

[0049] S2, constructing a patrol path inward based on the edge range of the indoor space, controlling the robot to move according to the patrol path and perform detour and obstacle avoidance through real-time distance measurement, and calibrating the actual moving path within the edge range of the indoor space to obtain a full path map; the detour and obstacle avoidance detours by maintaining a constant distance from the obstacle until returning to the patrol path;

[0050] S3. Based on the full path diagram, the obstacle area that the patrol line path cannot pass through is identified, and the edge of the obstacle area is rasterized to be divided into several sub-areas; a coding value is assigned to the sub-area according to the edge direction data of the obstacle area contained in the sub-area, and the coding value array is obtained by arranging the sub-areas according to the adjacent relationship, and the coding value array is traversed through the preset mean window to remove duplicates to obtain a feature value array;

[0051] S4, traversing the characteristic value array and the preset column characteristic value library to determine the column type;

[0052] S5. Construct an indoor space plan based on the edge range of the indoor space and the obstacle area of ​​the determined column type, map the column into the indoor space plan, and visualize the semantic information of the column.

[0053] In some embodiments, a mobile robot is first used to detour along the planar edge of an indoor building. The robot is equipped with a ranging sensor and a positioning module, which can accurately obtain its coordinate position and distance information from the surrounding environment during movement. In step S1, the robot collects its coordinate data along the edge of a wall or other building point by point, and uses a ranging sensor to continuously measure the distance from the wall, thereby constructing a preliminary edge contour of the indoor space. Throughout the process, the robot relies on real-time feedback of spatial coordinate information to avoid positioning errors caused by human intervention. In step S2, based on the indoor edge data obtained by the robot, the system automatically generates a patrol path through a preset path planning algorithm. The path is gradually advanced from the outside to the inside, ensuring global coverage while being able to efficiently identify all obstacles and structures in the indoor space. During the cruise, the robot detects obstacles in front in real time through a ranging sensor, and performs a detour operation to ensure a constant distance from the obstacle. The detour obstacle avoidance process relies on the robot's dynamic feedback mechanism, so that it can accurately return to the established patrol path after avoiding obstacles, avoiding path deviations during navigation. After the cruise is completed, the robot calibrates the actual moving path within the indoor edge range and generates a full path map containing the complete building edge and obstacles. In step S3, the system analyzes the generated full path map and calibrates the area where the patrol line path cannot pass, that is, the area where the obstacle is located. In order to improve the recognition accuracy, the edge data of the obstacle area is further rasterized. The system divides the edge of the obstacle into several sub-areas according to the predetermined grid units, and each sub-area contains part of the obstacle edge information. In order to accurately extract the spatial characteristics of the obstacle, the vector quantization method is used to analyze the edge direction data of each sub-area and assign it a corresponding encoding value. The encoding value represents the geometric direction and angle of the edge in the area. By analyzing the spatial position relationship of adjacent sub-areas, the system arranges these encoding values ​​in order to form an encoding value array. Next, the system processes the encoding value array in step S4. In order to reduce redundant information and improve recognition efficiency, the system traverses the encoding value array through a preset mean window, removes the encoding values ​​that are the same as the window mean, and finally obtains the feature value array after deduplication processing. The eigenvalue array is traversed and compared with the preset column eigenvalue library to determine whether the current obstacle area belongs to the column type. Through this eigenvalue comparison, the system can accurately identify the geometric shape and position of the column. In step S5, the system constructs a complete floor plan of the indoor space based on the identified column type and its spatial position, combined with the edge range of the indoor space. The floor plan not only contains the edge information of the building, but also accurately calibrates the main structures such as the column. Finally, the system maps the geometric information, type information and spatial coordinates of the column in the building into the floor plan, and displays its semantic information in a visual way, including the type, size and relative position of the column.In this way, the system generates a complete indoor space model, which can provide accurate spatial data support for subsequent architectural design, construction planning and management.

[0054] Furthermore, the robot includes a positioning module, a distance measurement module and a movement module;

[0055] The mobile module is used to control the steering and movement of the robot and record its parameters;

[0056] The distance measuring module is used to obtain the distance between the robot body and other objects through a laser distance measuring sensor;

[0057] The positioning module is used to calculate the robot displacement data through GPS to verify the parameters of steering and movement.

[0058] It should be noted that the mobile module is the core part of the robot's motion control. It achieves precise control of the robot's moving speed, direction and path through an integrated motor drive system and motion controller. When the robot performs detour or obstacle avoidance operations, the mobile module receives path planning instructions from the control system in real time and makes motion adjustments based on actual environmental feedback. For example, when the robot detects an obstacle in front of it, the mobile module can immediately perform a steering operation to make the robot detour along the specified obstacle avoidance path, and after the detour, automatically adjust the direction back to the original line patrol path. In addition, the mobile module also records the robot's movement parameters, such as driving distance, steering angle, etc. These parameters are not only used for subsequent data analysis, but also provide a basis for the correction of the robot's path and position recording. Secondly, the ranging module is implemented by a laser ranging sensor, which is responsible for obtaining the distance information between the robot and surrounding objects in real time. The laser ranging sensor uses a laser beam to measure the distance between the robot body and obstacles, walls or columns in the environment, ensuring that the robot can accurately perceive the obstacle position and boundary information in the space. When the robot cruises along the edge of the building, the ranging module can continuously collect distance data of edge objects, thereby helping the robot build a spatial model of the surrounding environment and respond to obstacles encountered in the path in a timely manner. For example, when the robot approaches a column or wall, the laser ranging sensor will inform the system of the distance change of the obstacle in real time through continuous data feedback, ensuring that the robot can adjust the detour path according to the ranging information while maintaining a constant safe distance from the obstacle. Finally, the positioning module can work together through the integrated GPS system and inertial navigation system (INS) to accurately measure and verify the robot's displacement data. Although the GPS signal may be subject to certain restrictions in indoor environments, in some semi-open or large buildings, the robot's position can still be comprehensively corrected through a combined navigation system (such as fusing inertial sensors and a small amount of GPS signals). The positioning module is not only used for initial positioning, but also for trajectory correction of the robot during movement. For example, when the robot is detouring or avoiding obstacles, the positioning module calculates the actual displacement data and compares it with the preset path information to ensure that the robot can travel accurately in the space. If a slight path deviation occurs during the detour, the positioning module can adjust the trajectory based on GPS data and inertial navigation feedback inside the robot to ensure that the robot's cruising path remains consistent with the preset trajectory.

[0059] Furthermore, the method of causing the robot to automatically move around the edge of an indoor plane includes the following steps:

[0060] Based on the SLAM algorithm, the edge contour information of the indoor space is obtained and the initial position of the robot is determined;

[0061] Use the A* algorithm to plan the optimal path for the robot to move along the edge of the room based on the edge contour, and adjust the path in real time to maintain a constant distance from the edge;

[0062] Based on the Hough transform algorithm, the geometric features of the indoor plane edge are extracted from the ranging data in real time, and the robot is controlled to avoid obstacles according to the ranging data;

[0063] The iterative closest point algorithm is used to correct the path after the robot completes the detour to close the path loop.

[0064] It should be noted that, first, when the robot starts, the SLAM (Simultaneous Localization and Mapping) algorithm is used to perform real-time positioning and map construction of the indoor environment. The core of the SLAM algorithm is that the robot can collect environmental data through sensors during movement, construct the edge contour information of the indoor environment in real time, and calculate the current coordinates of the robot at the same time. The SLAM algorithm combines laser ranging data with the robot's own movement information to establish a spatial coordinate system, so that the robot can accurately locate its position in the environment. This step is crucial to determine the initial position of the robot, especially in the building space, to ensure that the robot can perform the cruise task from the predetermined starting point. Next, based on the indoor edge contour information obtained by the SLAM algorithm, the robot uses the A* path planning algorithm to calculate the optimal path to move along the edge of the building. As a classic heuristic search algorithm, the A algorithm can plan the shortest and most efficient movement path under constraints such as obstacles and path length. In order to ensure that the robot maintains a constant distance from the edge of the indoor building during actual cruising, the system dynamically adjusts the path planning results. Specifically, the A* algorithm is not only used for initial path planning, but also updates the robot's position and distance information from the edge during the cruising process through a real-time feedback mechanism, ensuring that the robot always maintains a fixed safe distance from the wall or other building edges. For example, when the robot approaches the corner of a building, the A* algorithm dynamically adjusts the path by reanalyzing the environmental data, allowing the robot to turn smoothly and continue to move along the edge. In order to further extract the geometric features of the indoor plane edge, the robot relies on the Hough transform algorithm to process the edge data obtained from the ranging sensor during the cruising process. The Hough transform algorithm is particularly suitable for detecting edge features of geometric shapes such as straight lines and circles. The algorithm analyzes the point set in the laser ranging data, matches the measured points with the geometric contours in the environment, and extracts the edge segments of the building. The application of the Hough transform enables the robot to perceive the edge structures such as walls and columns around it in real time, and dynamically adjust the path based on these structural information. For example, when the robot detects the geometric edges of walls or columns during cruising, the Hough transform algorithm can quickly calculate the direction and position of these edges, allowing the robot to move accurately along these edges and bypass obstacles. When encountering an obstacle, the robot determines the specific location and size of the obstacle through the distance measurement data and the geometric features extracted by the Hough transform, and performs a detour. During this process, the system will analyze the relative position of the robot and the obstacle in real time to ensure that a constant safe distance is maintained from the obstacle during detour. When the detour is completed, the robot automatically returns to the original cruising path and continues to cruise along the edge. Finally, after the entire detour task is completed, the robot corrects the path through the iterative closest point (I CP) algorithm to ensure the closed-loop integrity of the cruising path.The ICP algorithm calculates the deviation between the current cruising path and the edge of the environment based on the environmental point cloud data collected by the robot. Through iterative calculation, the path deviation is minimized to ensure that the actual path of the robot is highly consistent with the preset cruising path. When the robot reaches the origin or the path closure position, the ICP algorithm can effectively correct the path deviation caused by measurement error or environmental change to ensure the closed-loop characteristics of the path.

[0065] Furthermore, obtaining the edge range of the indoor space according to the mobile coordinates and the ranging sensor includes:

[0066] Use the positioning module on the robot to obtain its current position, and determine the reference coordinates of the edge area in combination with the initial position indoors;

[0067] Obtain the distance data between the robot and the wall in real time and generate a set of distance measurement points at the indoor edge;

[0068] The least square fitting algorithm is used to process the data of the ranging point set to fit the geometric shape representing the indoor edge;

[0069] Based on the fitted edge geometry, the effective edge range of the indoor space is calculated and mapped to the indoor space coordinate system.

[0070] In some embodiments, first, when the robot starts, it obtains the current position through its internal positioning module. The positioning module is usually composed of a combination of multiple sensors, including a wheel odometer, which is used to accurately measure the displacement of the robot inside the building. The output data of the positioning module is matched with the initial position of the robot in the building to determine the starting position of the current robot and the preliminary motion reference system. This operation is the basis for subsequent path planning and edge recognition, ensuring that the robot can accurately track its own movement trajectory in the indoor environment. When the robot starts to cruise and move, the ranging sensor (such as a laser rangefinder or an ultrasonic rangefinder) begins to continuously obtain the distance information between the robot and the indoor wall. The ranging sensor generates a series of ranging points in real time by emitting a laser beam and calculating the time difference of the laser return. The collection of these ranging points is the distance data between the robot and the wall during the process of moving along the path. These data not only reflect the current environment of the robot, but also provide a basis for the next step of fitting the edge profile. In order to ensure the continuity and accuracy of the measurement, the sampling rate of the sensor is usually set to a high-frequency mode to ensure that sufficiently dense ranging data can still be obtained when the robot moves quickly. After completing the preliminary distance data collection, the system uses the least squares fitting algorithm to process the distance measurement point sets. Specifically, the least squares fitting is an optimization algorithm that calculates the geometric shape that best represents the measurement data by minimizing the error between the fitting curve and the measurement point. In this embodiment, the least squares fitting algorithm is used to extract the geometric shape features of indoor edges such as walls and columns. The algorithm fits the geometric contours such as straight lines and arcs of walls or other boundaries through iterative processing of distance data. This fitting method can effectively eliminate the interference of noise data and ensure that the geometric shape of indoor edges can be accurately extracted even in the case of large measurement errors. For example, when the robot moves along the indoor wall, the ranging sensor will continuously record the distance to the wall. As the robot moves along a straight path or a turning point, the geometric features of the distance measurement point set will show linear or curvilinear changes. Through the least squares fitting algorithm, the system can fit these point sets into straight line segments or arc segments, accurately outlining the actual edge shape of the indoor wall. Finally, based on the fitted edge geometry, the system further calculates the effective edge range of the entire indoor space. At this point, the fitted geometric shape is no longer a discrete distance measurement point, but an overall continuous boundary that reflects the edge information of indoor walls or other fixed objects. The system maps these fitted edge geometric shapes to the indoor spatial coordinate system to generate an indoor space model containing boundary information such as walls and columns. During this process, the system will perform high-precision calibration on geometric features such as wall corners and edge protrusions to ensure that the generated boundary lines are consistent with the actual spatial structure.

[0071] Furthermore, constructing a patrol path inwardly based on the edge range of the indoor space includes generating a patrol path based on the edge contour of the indoor space through a path planning algorithm, and the patrol path is constructed gradually inwardly from the edge in a round-trip form and covers the entire indoor plane space.

[0072] In some embodiments, when generating the initial patrol path, the system first determines the starting point position in the indoor plane, which is usually located in a corner or a straight line segment near the indoor edge. When the robot starts at the starting point, it starts to move along the edge with a certain initial direction. At this time, the system ensures that the robot always maintains a constant distance to follow the edge by calculating the distance between the robot and the edge. In this way, the robot will cruise along the entire edge step by step and record all the path points passed. The path planning of this step mainly depends on the information of the edge contour, and the continuity of the path points ensures the smooth progress of the edge cruise. In order to achieve full coverage cruise from the edge to the inside, the system will dynamically adjust the path after the initial edge path is completed to generate subsequent round-trip patrol paths. Specifically, when the robot completes a circle of edge path cruise, the system will advance the path from the outer edge to the inside according to the set interval distance. For example, the system can set a fixed interval distance (such as 0.5 meters or 1 meter) according to the cruise area or spatial form, so that the robot moves inward by a set distance after completing the outer circle path, and then starts the next circle of cruise. This inward advancement process is carried out layer by layer to ensure that the cruise covers the entire plane space. During each layer of the path, the system dynamically evaluates the spatial form to ensure that the path of each circle is parallel to the previous cruise path and the coverage interval remains uniform. During the path adjustment process, the system calculates the optimal path to avoid obstacles or complex terrain by combining the heuristic search function of the A* algorithm. For example, when the robot cruises to the corner of a wall or a column area, the system automatically calculates the best path to bypass these structures, ensuring that the robot returns to the original planned path smoothly while maintaining a safe distance from obstacles.

[0073] Furthermore, S3 comprises the following steps:

[0074] Based on the edge of the obstacle area in the full path graph, edge data is extracted by a boundary detection algorithm, and the edge of the obstacle area is divided into a plurality of rectangular grids according to a predetermined grid unit size, each of which has a unique index identifier;

[0075] For each rectangular grid, according to the edge direction information of the obstacle area contained therein, vector quantization is applied to divide the edge direction data into several angle intervals and set encoding values, wherein the edge direction data includes angle size and line direction;

[0076] According to the spatial position relationship of adjacent rectangular grids, an adjacency matrix between grid cells is constructed, and the code values ​​are arranged according to the spatial distribution order of the grid cells in the adjacent matrix to form a code value array;

[0077] Traversing the encoding value array through a preset first mean window, removing encoding values ​​in the preset mean window that are equal to the mean of the preset mean window, to obtain a semi-deduplicated array; the size of the preset first mean window is three encoding value units;

[0078] The encoding value array is traversed through the preset second mean window, and the encoding values ​​in the preset mean window that are equal to the mean of the preset mean window are removed to obtain the characteristic value array, and the size of the preset first mean window is two encoding value units.

[0079] In some embodiments, first, the system locates all impassable obstacle areas based on the full path map generated during the robot's cruising process. These areas usually include building structures such as columns and walls. In order to further process the edge information of these obstacles, the system applies a boundary detection algorithm to extract the edge data of the obstacle area. The boundary detection algorithm identifies the outline of the obstacle by analyzing the gradient changes at the edge of the obstacle. The algorithm can ensure that the shape of the obstacle area can be accurately outlined even in a complex environment, providing reliable basic data for subsequent processing. Next, the system rasterizes the extracted edge data and divides the obstacle area into several rectangular grids according to a predetermined grid unit size. The size of each rectangular grid is set by the algorithm according to the size, shape and spatial resolution of the obstacle. These rectangular grids provide a basic unit that is easy to manage for the block processing of the obstacle area. The system assigns a unique index identifier to each grid to ensure that each grid can be tracked and analyzed independently. In actual operation, the size of the grid unit needs to be adaptively adjusted according to the complexity and accuracy requirements of the obstacle to capture the subtle structural information in the edge. After completing the grid division, the system conducts an in-depth analysis of each rectangular grid to extract the edge direction information contained therein. To this end, the system applies vector quantization (VQ) technology to quantize the edge direction data contained in the grid. Specifically, vector quantization divides the direction data of the obstacle edge into several predetermined angle intervals and assigns a unique code value to each interval. The code value not only reflects the direction of the edge line in the grid, but also takes into account the geometric characteristics of the edge, such as the size of the angle and the direction of the line. For example, for a grid containing a 90-degree right angle, its code value will be different from that of a grid containing only linear edges. Through this encoding process, the system effectively converts complex edge data into standardized discrete data. Next, the system constructs an adjacency matrix based on the spatial position relationship of each grid cell, which represents the spatial topological relationship between adjacent grids. Based on the adjacency matrix, the system arranges all the code values ​​in order according to the spatial distribution order of adjacent grid cells to form a code value array. This array not only contains the direction information of each grid, but also reflects the relative position of the grid in space, ensuring that the code value can accurately describe the overall outline and structure of the obstacle area. After the code value array is generated, the system traverses the array through the preset first mean window. The mean window is a sliding window through which the system calculates the mean of the code values ​​and removes the code values ​​that are equal to the window mean. In the initial processing of the code value array, a larger window of three code value units is first used, mainly to retain the overall characteristic information of the data during the deduplication process while reducing redundancy.In the initial code value array, there may be multiple consecutive identical or similar code values, most of which come from adjacent grids, reflecting the consistency in the edge direction. By setting a relatively large window (three units), the system can calculate a more representative mean and remove the code values ​​equal to the window mean. The purpose of this step is to preliminarily eliminate redundant information while retaining the overall geometric structure features. For example, if a part of the edge of an obstacle continuously presents the same direction information (such as a long straight line), the redundant information in this long straight line can be effectively removed through a window of three code value units, leaving only one representative code value. Therefore, the first mean window is set to three units, ensuring that the deduplication operation can filter out duplicate information while retaining the main features, thereby reducing the amount of data to be processed subsequently. The second mean window size is set to two code value units, mainly to refine the processing of the remaining code values. After the initial deduplication of the first window, the redundant information in the code value array has been greatly reduced, but there may still be some small redundant or locally similar code values ​​on some edge structures. At this point, the system needs to further refine the remaining array of coded values ​​to remove some eigenvalues ​​that are still too similar in local areas. Since the amount of data has been reduced at this stage, the system can use a smaller window (two units) to achieve more accurate deduplication. A two-unit window can more keenly capture changes within a short distance, such as smaller angle changes or shorter edge turns. By deduplicating the mean of a smaller window, the system is able to retain these subtle structural features without ignoring important local features due to an overly large window.

[0080] Further, the S4 comprises the following steps:

[0081] Perform equivalent sorting according to the eigenvalue array to generate several equivalent eigenvalue sequences;

[0082] The preset columnar feature value library is compared with the equivalent feature value sequence and the similarity value is calculated. If the similarity value is lower than the preset similarity threshold, no processing is performed;

[0083] If the similarity value is greater than or equal to a preset similarity threshold, the obstacle area is marked as a column type described by a preset column feature value library.

[0084] In some embodiments, first, after obtaining the eigenvalue array of the obstacle area, the system performs equivalent sorting on the array. The so-called equivalent sorting means that the content of the eigenvalue array may be different arrangements of the same column structure, that is, different orders are generated by rotation or changes in the arrangement method, while its essential information remains unchanged. For example, if the eigenvalue array of the column is (1,2,3,4), multiple equivalent sequences can be generated through equivalent sorting, such as (2,3,4,1), (3,4,1,2), (4,1,2,3) and (1,2,3,4). The generation of this equivalent sorting ensures that no matter how the arrangement form of the eigenvalue array changes, the system can identify the same geometric features corresponding to them. In actual operation, the system generates all equivalent eigenvalue sequences by traversing each possible arrangement and combination. These sequences achieve different arrangements only by rotation or movement while keeping the original eigenvalue unchanged. This equivalent sorting process is particularly suitable for columns with strong symmetry or possible diverse arrangement orders, ensuring that the system can fully cover all possible eigenvalue combinations to avoid omissions or misjudgments. Next, the system traverses and compares the generated equivalent feature value sequence with the preset column feature value library. The column feature value library is a pre-set database that contains standard feature value arrays of different types of columns and their equivalent forms. The system calculates the similarity by comparing the feature value arrays with the standard features in the feature value library one by one. The calculation of the similarity value is based on a distance measurement algorithm, such as Euclidean distance or cosine similarity. The algorithm evaluates the degree of difference between the feature value array to be identified and the standard features in the feature value library to derive a similarity value to measure the similarity between the two. During the comparison process, the system calculates the similarity of each equivalent feature value sequence and compares the calculation result with the preset similarity threshold. The similarity threshold is a measurement standard set by the system to distinguish whether the degree of match is high enough. If the similarity value of a feature value sequence is lower than the threshold, it means that the sequence does not match the column feature in the feature value library, and the system will not further process the sequence to avoid misjudgment. This can effectively filter out obstacle information that is not related to the column feature. If the system finds that the similarity value of a certain equivalent eigenvalue sequence is greater than or equal to the preset similarity threshold, it indicates that the sequence has a high similarity with a standard column feature in the eigenvalue library. In this case, the system determines that the current obstacle area is likely to belong to a preset column type. The system will mark the obstacle area as the column type described by the preset column eigenvalue library, and store its information in the spatial model for subsequent visualization and processing.

[0085] Furthermore, the formula for the equivalent sorting is as follows:

[0086]

[0087] Among them, S is the eigenvalue array, containing n eigenvalues; k is the number of rotations; i is the index; S k It is a new array obtained by rotating the eigenvalue array S k times, that is, the rotated eigenvalue sequence; mod is the modular operation.

[0088] Specifically, the formula can be understood as: starting from an eigenvalue array, each time a new arrangement is generated by rotating the elements in the array. Regardless of how the elements of the original array are arranged, all possible arrangements can be generated by this formula. Eigenvalue array S: This is an array that represents the characteristics of the cylinder, containing a series of data describing the geometric shape of the cylinder, such as edge direction, angle size, etc. Assume that the eigenvalue array is S = [1,2,3,4], where n = 4 means that there are 4 elements in the array. The formula generates multiple equivalent arrangements through rotation operations. Each rotation is equivalent to moving the elements in the array to the right. For example, during the first rotation, the array changes from S = [1,2,3,4] to S1 = [2,3,4,1], that is, the first element moves to the end and the remaining elements move forward. After the second rotation, the array becomes S2 = [3,4,1,2], and so on. Modulo operations (i.e., remainder operations) ensure that the array does not exceed the range of the index when it is rotated. When the number of rotations exceeds the length of the array, the modulo operation will reposition the index to the beginning of the array. For example, when the array length is 4 and the number of rotations is 5, the modular operation will cause the rotated array to recalculate the arrangement from the beginning. Although these arrangements are in different orders, they actually represent the same geometric features. By generating all possible arrangements, the system can handle different orders of the eigenvalue array and ensure that the system can identify the same type of cylinder regardless of the arrangement input. Eliminate the influence of arrangement order: In practical applications, the characteristics of the cylinder may cause the arrangement of the eigenvalues ​​to change due to different measurement angles or orders. Through the equivalent sorting formula, the system can ignore these differences in arrangement and identify different arrangements as the same cylinder features.

[0089] Further, the S5 comprises the following steps:

[0090] Based on the edge range and full path graph of the indoor space, the indoor space plan is constructed using the GIS algorithm;

[0091] According to the obstacle area corresponding to the column, its coordinates and range in the plan view are determined and mapped in the indoor space plan view;

[0092] Different types of columns are semantically labeled using identifiers.

[0093] In some embodiments, first, the system uses the GIS algorithm to construct a complete indoor space plan based on the indoor space edge range and full path map obtained in the previous step. The GIS algorithm can effectively process spatial data, especially modeling the geometric structure of buildings in two-dimensional or three-dimensional space. The system analyzes the edge data collected during the robot's cruising process, maps these data to a standard indoor coordinate system, and draws the edge information of the building on the plan. This step ensures that the structure of the entire indoor environment is accurately digitally represented to form a visual plan map. Next, the system locates the coordinates and range of each column in the indoor space plan based on the characteristic value recognition results of the column. In the previous step, the obstacle area corresponding to the column has been identified as a specific column type. Now the system accurately maps the geometric boundaries of the obstacle area (such as the center position and radius of the column) to the indoor space plan. To this end, the system calculates the specific spatial coordinates and coverage of each column based on the relative position of each column in the full path map and its distance from the indoor edge. This step ensures the accurate positioning of the column in the spatial plan and provides basic data for subsequent visualization operations. After the coordinates and range of the column are determined, the system will semantically annotate different types of columns through identifiers. The identifier can be a code, name, or other morphological information identification of the column type, such as "round column", "square column" or "T column". Through these identifiers, the system can distinguish and mark different columns on the floor plan, ensuring that the user or system can quickly identify the specific type of each column and its function in the space. These annotations can not only be used for visual display, but also for subsequent architectural design, space management or maintenance planning. In this way, the system can build a detailed indoor space model containing the location, type and other semantic information of the column. This semantic annotation not only makes it easier for later users to view and understand the specific type of the column, but also provides important reference information for the subsequent design and construction of the building. In addition, semantic annotation can also be used for classification management or maintenance of different column types to ensure that the structural support elements of the entire building can be effectively monitored and managed during future use.

[0094] The above implementation modes are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary engineering and technical personnel in the field shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A column positioning and identification method based on spatial information, characterized in that: The following steps are involved: S1, making the robot automatically move around the edge of the indoor plane, and obtaining the edge range of the indoor space according to the moving coordinates and ranging sensors; S2, constructing a patrol path inward based on the edge range of the indoor space, controlling the robot to move according to the patrol path and perform detour and obstacle avoidance through real-time distance measurement, and calibrating the actual moving path within the edge range of the indoor space to obtain a full path map; the detour and obstacle avoidance detours by maintaining a constant distance from the obstacle until returning to the patrol path; S3. Based on the full path diagram, the obstacle area that the patrol line path cannot pass through is identified, and the edge of the obstacle area is rasterized to be divided into several sub-areas; a coding value is assigned to the sub-area according to the edge direction data of the obstacle area contained in the sub-area, and the coding value array is obtained by arranging the sub-areas according to the adjacent relationship, and the coding value array is traversed through the preset mean window to remove duplicates to obtain a feature value array; S4, traversing the characteristic value array and the preset column characteristic value library to determine the column type; S5. Construct an indoor space plan based on the edge range of the indoor space and the obstacle area of ​​the determined column type, map the column into the indoor space plan, and visualize the semantic information of the column.

2. A column positioning and identification method based on spatial information according to claim 1, characterized in that: The robot comprises a positioning module, a distance measuring module and a moving module; The mobile module is used to control the steering and movement of the robot and record its parameters; The distance measuring module is used to obtain the distance between the robot body and other objects through a laser distance measuring sensor; The positioning module is used to calculate the robot displacement data through GPS to verify the parameters of steering and movement.

3. A column positioning and identification method based on spatial information according to claim 1, characterized in that: The method of causing the robot to automatically move around along the edge of an indoor plane comprises the following steps: Based on the SLAM algorithm, the edge contour information of the indoor space is obtained and the initial position of the robot is determined; Use the A* algorithm to plan the optimal path for the robot to move along the edge of the room based on the edge contour, and adjust the path in real time to maintain a constant distance from the edge; Based on the Hough transform algorithm, the geometric features of the indoor plane edge are extracted from the ranging data in real time, and the robot is controlled to avoid obstacles according to the ranging data; The iterative closest point algorithm is used to correct the path after the robot completes the detour to close the path loop.

4. The column positioning and identification method based on spatial information according to claim 1, characterized in that: The step of obtaining the edge range of the indoor space according to the mobile coordinates and the ranging sensor includes: Use the positioning module on the robot to obtain its current position, and determine the reference coordinates of the edge area in combination with the initial position indoors; Obtain the distance data between the robot and the wall in real time and generate a set of distance measurement points at the indoor edge; The least square fitting algorithm is used to process the data of the ranging point set to fit the geometric shape representing the indoor edge; Based on the fitted edge geometry, the effective edge range of the indoor space is calculated and mapped to the indoor space coordinate system.

5. A column positioning and identification method based on spatial information according to claim 4, characterized in that: The constructing of the patrol path inwardly based on the edge range of the indoor space includes generating the patrol path based on the edge contour of the indoor space through a path planning algorithm, and the patrol path is constructed gradually inwardly from the edge in a round-trip form and covers the entire indoor plane space.

6. The column positioning and identification method based on spatial information according to claim 1, characterized in that: The S3 comprises the following steps: Based on the edge of the obstacle area in the full path graph, edge data is extracted by a boundary detection algorithm, and the edge of the obstacle area is divided into a plurality of rectangular grids according to a predetermined grid unit size, each of which has a unique index identifier; For each rectangular grid, according to the edge direction information of the obstacle area contained therein, vector quantization is applied to divide the edge direction data into several angle intervals and set encoding values, wherein the edge direction data includes angle size and line direction; According to the spatial position relationship of adjacent rectangular grids, an adjacency matrix between grid cells is constructed, and the code values ​​are arranged according to the spatial distribution order of the grid cells in the adjacent matrix to form a code value array; Traversing the encoding value array through a preset first mean window, removing encoding values ​​in the preset mean window that are equal to the mean of the preset mean window, to obtain a semi-deduplicated array; the size of the preset first mean window is three encoding value units; The encoding value array is traversed through the preset second mean window, and the encoding values ​​in the preset mean window that are equal to the mean of the preset mean window are removed to obtain the characteristic value array, and the size of the preset first mean window is two encoding value units.

7. A column positioning and identification method based on spatial information according to claim 6, characterized in that: The S4 comprises the following steps: Perform equivalent sorting according to the eigenvalue array to generate several equivalent eigenvalue sequences; The preset columnar feature value library is compared with the equivalent feature value sequence and the similarity value is calculated. If the similarity value is lower than the preset similarity threshold, no processing is performed; If the similarity value is greater than or equal to a preset similarity threshold, the obstacle area is marked as a column type described by a preset column feature value library.

8. A column positioning and identification method based on spatial information according to claim 7, characterized in that: The formula for the equivalent sorting is as follows: Among them, S is the eigenvalue array, containing n eigenvalues; k is the number of rotations; i is the index; S k It is a new array obtained by rotating the eigenvalue array S k times, that is, the rotated eigenvalue sequence; mod is the modular operation.

9. The column positioning and identification method based on spatial information according to claim 1, characterized in that: The S5 comprises the following steps: Based on the edge range and full path map of the indoor space, the indoor space plan is constructed using GIS algorithm; According to the obstacle area corresponding to the column, its coordinates and range in the plan view are determined and mapped in the indoor space plan view; Different types of columns are semantically labeled using identifiers.

Images