Method for measuring space distance between dangerous source and target by using monocular vision

Abstract

The invention discloses a method for measuring the space distance between a dangerous source and a target by using monocular vision, and relates to the related field of distance measurement, and the method comprises the steps: carrying out the deployment of front-end collection equipment; the laser projection device is controlled to carry out target scene global scanning, laser point cloud data are transmitted back, the visual camera is controlled to carry out scene full-field collection, and image data are transmitted back; constructing a scene space model according to the laser point cloud data and the image data; and determining a target point cloud position and performing spatial distance measurement through the scene space model based on matching of an image space and a point cloud two-dimensional projection space and point cloud mapping of the point cloud two-dimensional projection space and a point cloud three-dimensional space, and determining a target spatial distance. According to the invention, the technical problem of insufficient precision of the existing monocular vision-based dangerous source and target space distance measurement is solved, and the technical effect of improving the space distance measurement precision is achieved.

Description

Technical Field

[0001] This application relates to the field of distance measurement, and particularly to a method for measuring the spatial distance between a hazard source and a target using monocular vision. Background Art

[0002] In the operation and maintenance and monitoring of transmission lines, accurately measuring the spatial distance between a hazard source and a transmission line helps to evaluate potential safety hazards. Currently, the common practice is to attempt to extract depth information from a single image through complex image processing algorithms and geometric transformations, and then match it with three-dimensional point cloud data. However, in practical applications, due to the influence of factors such as lighting conditions, object surface texture, and occlusion, a single image cannot directly provide sufficient depth information, resulting in parallax errors when converting the two-dimensional target position to three-dimensional coordinates, thereby affecting the accuracy of distance measurement.

[0003] In the current related technologies, there are technical problems of insufficient accuracy in measuring the spatial distance between a hazard source and a target using monocular vision. Summary of the Invention

[0004] This application provides a method for measuring the spatial distance between a hazard source and a target using monocular vision. By scanning the target scene to obtain laser point cloud data, taking a single image to obtain image data, combining the laser point cloud and the image data, establishing a scene space model including a three-dimensional space, a two-dimensional projection space, and an image space, performing matching between the image space and the point cloud two-dimensional projection space in the model, and then mapping from the point cloud two-dimensional projection space to the point cloud three-dimensional space to determine the target point cloud position, and based on the determined target point cloud position, performing spatial distance measurement to obtain the target spatial distance and other technical means, achieving the technical effect of improving the accuracy of spatial distance measurement.

[0005] This application provides a method for measuring the spatial distance between a hazard source and a target using monocular vision, including: deploying front-end acquisition devices, where the front-end acquisition devices include a laser projection device and a vision camera; controlling the laser projection device to perform a full-field scan of the target scene and transmit back the laser point cloud data, and controlling the vision camera to perform a full-field acquisition of the scene and transmit back the image data, where the image data is a single image; constructing a scene space model with the laser point cloud data and the image data, where the scene space model includes a point cloud three-dimensional space - a point cloud two-dimensional projection space - an image space; through the scene space model, based on the matching of the image space - the point cloud two-dimensional projection space, and the point cloud mapping of the point cloud two-dimensional projection space - the point cloud three-dimensional space, determining the target point cloud position and performing spatial distance measurement to determine the target spatial distance, where the target point cloud position is the point cloud position of the hazard source and the dangerous target.

[0006] In a possible implementation, for constructing the scene space model, the following processing is performed: traversing the lidar point cloud data to identify the point cloud coordinate data and scene feature data of each lidar point cloud; based on the point cloud coordinate data, building the three-dimensional point cloud space, where the three-dimensional point cloud space represents the spatial position of the point cloud; combining the scene feature data, and taking the three-dimensional point cloud space as a reference to construct a two-dimensional point cloud projection space, where the two-dimensional point cloud projection space uses the field of view coordinate system of the image data as the projection plane; building an image space with the image data, hierarchically integrating the three-dimensional point cloud space - two-dimensional point cloud projection space - image space, and determining the scene space model.

[0007] In a possible implementation, for combining the scene feature data and taking the three-dimensional point cloud space as a reference to construct a two-dimensional point cloud projection space, the following processing is performed: determining the field of view coordinate system of the image data based on the acquisition parameters of the vision camera; constructing the projection plane according to the field of view coordinate system, establishing a point cloud projection from the three-dimensional point cloud space to the projection plane, and determining the two-dimensional point cloud projection; traversing the scene feature data to determine the feature descriptors of each lidar point cloud, and performing projection point cloud identification on the two-dimensional point cloud projection to generate the two-dimensional point cloud projection space.

[0008] In a possible implementation, for determining the position of the target point cloud, the following processing is performed: framing the hazard source and the dangerous target based on the image space; for the hazard source and the dangerous target, performing matching based on the two-dimensional point cloud projection space to determine the projection matching result, which includes position matching and feature matching; using the point cloud mapping between the two-dimensional point cloud projection space and the three-dimensional point cloud space to determine the three-dimensional mapped point cloud of the projection matching result as the position of the target point cloud.

[0009] In a possible implementation, for framing the hazard source and the dangerous target, the following processing is performed: for the target scene, mining the risk database based on historical risk control data; transmitting back the image data and performing filtering and noise reduction to determine the preprocessed image; traversing the risk database to perform matching on the preprocessed image to locate the hazard source and the dangerous target.

[0010] In a possible implementation, for performing matching based on the two-dimensional point cloud projection space, the following processing is performed: framing the hazard source based on the image space and determining the first coordinate, framing the dangerous target and determining the second coordinate; performing coordinate positioning on the two-dimensional point cloud projection space according to the first coordinate and the second coordinate to determine the first positioning coordinate; identifying the pixel features of the hazard source and the dangerous target, performing feature matching based on the feature descriptors to determine the second matching coordinate; proofreading and integrating the first positioning coordinate and the second matching coordinate to determine the projection matching result.

[0011] In a possible implementation, for the spatial distance measurement, the following processing is performed: identifying the position of the target point cloud, determining a first point cloud range and a second point cloud range, where the first point cloud range is the point cloud distribution of the hazard source, and the second point cloud range is the point cloud distribution of the dangerous target; based on the first point cloud range and the second point cloud range, determining a spatial distance measurement point cloud pair, where the spatial distance measurement point cloud is a boundary point cloud pair; measuring the relative distance of the spatial distance measurement point cloud pair to determine the target spatial distance, where the relative distance includes a straight-line distance, a horizontal distance, and a vertical distance.

[0012] In a possible implementation, for the determination of the spatial distance measurement point cloud pair, the following processing is performed: determining the central point cloud of the first point cloud range and the second point cloud range, and using the boundary point clouds on the line connecting the central point clouds as the first boundary point cloud and the second boundary point cloud; using the first boundary point cloud as a reference point, performing a boundary extension measurement based on the reference point to determine the first effective point cloud of the hazard source, where the boundary geometric features are determined based on the point cloud normal vector and curvature; using the second boundary point cloud as a reference point, performing a boundary extension measurement based on the reference point to determine the second effective point cloud of the dangerous target; and using the first effective point cloud and the second effective point cloud as the boundary point cloud pair.

[0013] It is intended to use the method for measuring the spatial distance between a hazard source and a target using monocular vision proposed in this application. First, the front-end acquisition device is deployed. The front-end acquisition device includes a laser projection device and a vision camera. Then, the laser projection device is controlled to perform a full-field scan of the target scene, and the laser point cloud data is transmitted back. The vision camera is controlled to perform a full-field acquisition of the scene, and the image data is transmitted back. The image data is a single image. Then, a scene space model is constructed based on the laser point cloud data and the image data. The scene space model includes a point cloud three-dimensional space - a point cloud two-dimensional projection space - an image space. Finally, through the scene space model, based on the matching of the image space - the point cloud two-dimensional projection space and the point cloud mapping of the point cloud two-dimensional projection space - the point cloud three-dimensional space, the position of the target point cloud is determined and the spatial distance is measured to determine the target spatial distance, where the position of the target point cloud is the point cloud positions of the hazard source and the dangerous target, achieving the technical effect of improving the accuracy of spatial distance measurement. Description of the Drawings

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments of the present invention will be briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the methods according to the embodiments of the present application. It should be understood that the operations before or below do not necessarily need to be executed precisely in sequence. On the contrary, according to the need, they can be executed in reverse order or simultaneously. At the same time, other operations can also be added to these processes, or one or several operations can be removed from these processes.

[0015] Figure 1 This is a flowchart showing the method for measuring the spatial distance between a hazard source and a target using monocular vision provided by an embodiment of the present application.

[0016] Figure 2 This is a flowchart showing the process of constructing a scene space model in the method for measuring the spatial distance between a hazard source and a target using monocular vision provided by an embodiment of the present application. Detailed implementation manners

[0017] The above description is only an overview of the technical solution of the present application. In order to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the following specifically gives the detailed implementation manners of the present application.

[0018] In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on the present application. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present application.

[0019] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict. The terms "first\second" involved are only used to distinguish similar objects and do not represent a specific order for the objects. The terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server including a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or modules not clearly listed or inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application.

[0020] An embodiment of the present application provides a method for measuring the spatial distance between a hazard source and a target using monocular vision, as Figure 1 shown, the method includes: Step S100, deploy the front-end acquisition device, where the front-end acquisition device includes a laser projection device and a vision camera.

[0021] Specifically, front-end acquisition devices are deployed, including laser projection devices and vision cameras. Among them, the laser projection device is a device that can emit a laser beam and receive the reflected laser signal, and is used to generate laser point cloud data. The vision camera is used to capture the image data of the target scene. A single-frame image camera is used, which can capture the two-dimensional image of the scene.

[0022] Step S200: Control the laser projection device to perform a full-field scan of the target scene and transmit back the laser point cloud data, and control the vision camera to perform a full-field acquisition of the scene and transmit back the image data, where the image data is a single-frame image.

[0023] Specifically, control the laser projection device to emit a laser beam to perform a full-field scan of the target scene. The laser beam will be reflected back during the scan to form laser point cloud data. These data contain the three-dimensional coordinate information of each point in the scene. While the laser projection device is scanning, control the vision camera to perform a full-field acquisition of the scene and capture the two-dimensional image data of the scene.

[0024] Step S300: Construct a scene space model with the laser point cloud data and the image data, where the scene space model includes a point cloud three-dimensional space - a point cloud two-dimensional projection space - an image space.

[0025] Specifically, use the laser point cloud data and the image data to construct a scene space model through means such as image matching, feature point extraction and matching. The scene space model is deployed on the front-end intelligent monitoring device, and the intelligent monitoring device can perform the identification and ranging processing of potential hazards. The scene space model is a model that includes the corresponding relationship between the point cloud three-dimensional space, the point cloud two-dimensional projection space and the image space. Among them, the point cloud three-dimensional space is a three-dimensional space composed of laser point cloud data, which contains the three-dimensional coordinate information of each point in the scene. The point cloud two-dimensional projection space is a space formed by projecting the points in the point cloud three-dimensional space onto a two-dimensional plane. This projection space is a bridge connecting the point cloud three-dimensional space and the image space. The image space is a space composed of the image data captured by the vision camera, which contains the two-dimensional image information of the scene.

[0026] Such as Figure 2As shown, in a possible implementation, for the step of constructing the scene space model, step S300 further includes step S310 of traversing the laser point cloud data to identify the point cloud coordinate data and the scene feature data of each laser point cloud. Specifically, the laser point cloud data transmitted back by the laser projection device is processed point by point through programming methods (such as using languages like Python, C++). The point cloud data includes the three-dimensional coordinates (X, Y, Z) and intensity information of each laser point. During the traversal process, the three-dimensional coordinates of each laser point are identified, which are the basis for constructing the three-dimensional point cloud space. At the same time, according to features such as the density, shape, and color of the point cloud, specific objects or feature regions in the scene are identified.

[0027] Step S320 of building the three-dimensional point cloud space based on the point cloud coordinate data, where the three-dimensional point cloud space represents the spatial position of the point cloud. Specifically, using a three-dimensional modeling software or programming library (such as PCL, Open3D, etc.), according to the point cloud coordinate data identified in step S310, a three-dimensional point cloud model of the target scene is constructed. This model is used to represent the three-dimensional structure and shape of the scene and is a set composed of countless three-dimensional points, and each point represents a specific position in the scene.

[0028] Step S330 of constructing a two-dimensional projection space of the point cloud with the three-dimensional point cloud space as the reference in combination with the scene feature data, where the two-dimensional projection space of the point cloud uses the field-of-view coordinate system of the image data as the projection plane. Specifically, according to the field-of-view coordinate system of the image data captured by the vision camera, a two-dimensional projection plane is determined, and this projection plane is the mapping plane from the three-dimensional point cloud space to the image space. Then, according to the point cloud coordinates and scene feature data in the three-dimensional point cloud space, these points are mapped to the projection plane according to projection rules (such as orthographic projection, perspective projection, etc.) to form the two-dimensional projection space of the point cloud. This space retains the three-dimensional position information of the point cloud but represents it in a two-dimensional form.

[0029] Step S340: Build an image space with the image data, hierarchically integrate the point cloud three-dimensional space - point cloud two-dimensional projection space - image space, and determine the scene space model. Specifically, use an image processing software or programming library (such as OpenCV, etc.) to construct an image space based on the image data captured by the vision camera. This space is a two-dimensional space that represents the pixel information and color information in the image. Integrate the point cloud three-dimensional space, point cloud two-dimensional projection space, and image space together according to the hierarchical relationship to form a complete scene space model. This model can reflect the three-dimensional structure, shape, and feature information in the image of the target scene. After the integration is completed, save the scene space model through programming or modeling software for applications such as spatial distance measurement and hazard source identification. This implementation method can map the target position in the image to the projection plane by introducing the point cloud two-dimensional projection space, thus avoiding the three-dimensional ranging error caused by the perspective difference.

[0030] In a possible implementation, in combination with the scene feature data, with the point cloud three-dimensional space as the reference, construct the point cloud two-dimensional projection space. Step S330 further includes step S331: Determine the field of view coordinate system of the image data based on the acquisition parameters of the vision camera. Specifically, the acquisition parameters of the vision camera include the focal length, optical center position, pixel size, etc. of the camera, which determine the corresponding relationship between the image captured by the camera and the actual scene. According to the acquisition parameters of the camera, calculate the field of view coordinate system of the image. This coordinate system is a two-dimensional coordinate system used to represent the corresponding relationship between the pixel positions in the image and the spatial positions in the actual scene.

[0031] Step S332: According to the field of view coordinate system, construct the projection plane, establish the point cloud projection from the point cloud three-dimensional space to the projection plane, and determine the point cloud two-dimensional projection. Specifically, construct the projection plane. The projection plane is a two-dimensional plane used to receive the points projected from the point cloud three-dimensional space. This plane corresponds to the field of view coordinate system of the image, that is, each point on the projection plane corresponds to a pixel in the image. Project the points in the point cloud three-dimensional space onto the projection plane according to the projection rules to form the point cloud two-dimensional projection. After the projection is completed, each point on the projection plane corresponds to a point in the point cloud three-dimensional space, but at this time these points have been converted into two-dimensional coordinates.

[0032] Step S333: Traverse the scene feature data to determine the feature descriptors of each laser point cloud, perform projection point cloud identification on the two-dimensional projection of the point cloud, and generate the two-dimensional projection space of the point cloud. Specifically, traverse the scene feature data, which includes information such as the shape, color, and texture of the point cloud. Determine a feature descriptor for each laser point cloud, which is used to uniquely identify the point cloud. The feature descriptor can be a combination of one or more of the shape features, color features, texture features, etc. of the point cloud. Through feature matching algorithms such as SIFT and SURF, associate the feature descriptor of each laser point cloud with its corresponding point in the two-dimensional projection of the point cloud to form the projection point cloud identification. After completing the projection point cloud identification, a space containing the positions of all laser point clouds on the two-dimensional projection plane and their feature descriptors is obtained, that is, the two-dimensional projection space of the point cloud. This space contains both the position information of the point cloud and their feature information, and can be used for matching with the image pixel features. This implementation method projects the point cloud in the three-dimensional space onto the two-dimensional plane by constructing the two-dimensional projection space of the point cloud while retaining their feature information, so that image processing algorithms can be used for feature matching on the two-dimensional plane, thereby improving the accuracy of hazard source identification and spatial distance measurement.

[0033] Step S400: Through the scene space model, based on the matching of the image space - two-dimensional projection space of the point cloud and the point cloud mapping of the two-dimensional projection space of the point cloud - three-dimensional space of the point cloud, determine the position of the target point cloud and perform spatial distance measurement to determine the target spatial distance, where the target point cloud position is the point cloud position of the hazard source and the dangerous target.

[0034] Specifically, first, identify the image feature points of the hazard source and the dangerous target in the image space. Then, use the corresponding relationship in the scene space model to map these feature points to the two-dimensional projection space of the point cloud. Next, use the corresponding relationship between the two-dimensional projection space of the point cloud and the three-dimensional space of the point cloud to further map the feature points mapped to the two-dimensional projection space of the point cloud to the three-dimensional space of the point cloud, so as to determine the point cloud positions of the hazard source and the dangerous target. Finally, calculate the spatial distance between them according to the point cloud positions of the hazard source and the dangerous target in the three-dimensional space of the point cloud. The embodiment of the present application uses technical means such as scanning the target scene to obtain laser point cloud data, taking a single image to obtain image data, combining the laser point cloud and the image data, establishing a scene space model including a three-dimensional space, a two-dimensional projection space, and an image space, performing matching between the image space and the two-dimensional projection space of the point cloud in the model, then mapping through the two-dimensional projection space of the point cloud to the three-dimensional space of the point cloud, determining the position of the target point cloud, and performing spatial distance measurement according to the determined position of the target point cloud to obtain the target spatial distance, achieving the technical effect of improving the accuracy of spatial distance measurement.

[0035] In a possible implementation, the step of determining the position of the target point cloud, step S400 further includes step S410 of framing the hazard sources and hazard targets based on the image space. Specifically, the image space is a space composed of image data collected by a vision camera, which contains two-dimensional pixel information of the scene. In the image space, image processing techniques (such as object detection, etc.) are used to identify and frame the positions of the hazard sources and hazard targets. Specifically, it includes preprocessing the image (such as denoising, enhancing contrast, etc.), and then applying object detection algorithms (such as convolutional neural networks, support vector machines, etc.) to identify the target objects and mark them with rectangular frames or other shapes in the image. Among them, the hazard source is the source that may cause danger or accidents, such as tree growth, construction activities, etc. The hazard target is the object that needs to be monitored or protected, such as the transmission line and its surrounding key areas.

[0036] Step S420, for the hazard sources and the hazard targets, perform matching based on the two-dimensional projection space of the point cloud to determine the projection matching result, which includes position matching and feature matching. Specifically, the two-dimensional projection space of the point cloud is a space composed of the projection of the laser point cloud data in the image two-dimensional coordinate system, which contains two-dimensional projection information of the scene. In the two-dimensional projection space of the point cloud, feature matching algorithms are used to find the projection points corresponding to the hazard sources and hazard targets framed in the image space. Specifically, it includes extracting the feature descriptors of the projection points (such as shape, color, texture, etc.), and then performing feature matching in the projection space to find the most similar projection points. After the matching is completed, a projection matching result including position matching and feature matching is obtained. Position matching means that the position of the projection point in the projection space corresponds to the target position framed in the image space; feature matching means that the feature descriptors of the projection point match the feature descriptors of the target object in the image space.

[0037] Step S430: Determine the three-dimensional mapped point cloud of the projection matching result as the target point cloud position based on the point cloud mapping between the two-dimensional projection space of the point cloud and the three-dimensional space of the point cloud. Specifically, the point cloud mapping between the two-dimensional projection space of the point cloud and the three-dimensional space of the point cloud establishes a mapping relationship between the two spaces, enabling each point in the projection space to find its corresponding point in the three-dimensional space. Using the point cloud mapping relationship, map the projection points in the projection matching result to the three-dimensional space of the point cloud to obtain the corresponding three-dimensional point cloud. These three-dimensional point clouds are the target point cloud positions, representing the actual positions of the hazard sources and dangerous targets in the three-dimensional space. This implementation method combines laser point cloud data and image data by constructing a scene space model. By using the matching relationship between the two-dimensional projection space of the point cloud and the image space, rapid identification and positioning of hazard sources and dangerous targets can be achieved. Then, through the point cloud mapping relationship, map the projection matching result to the three-dimensional space of the point cloud to obtain the target point cloud position, improving the accuracy of determining the positions of hazard sources and dangerous targets in the three-dimensional space.

[0038] In a possible implementation, for the step of framing the hazard sources and dangerous targets, step S410 further includes step S411: Mine the risk database based on historical risk control data for the target scene. Specifically, collect historical risk control data from various sources. The historical risk control data includes past line fault records, inspection reports, meteorological data, geographical information, etc., which contain various factors leading to line faults. Perform preprocessing operations such as data cleaning, duplicate removal, and format unification on the collected data to improve the efficiency and quality of data mining. Use feature extraction algorithms to extract features related to line faults from the historical data, such as fault types, fault locations, environmental factors, etc. Integrate the extracted feature information into the risk database to form a structured knowledge base.

[0039] Step S412: Transmit the image data back and perform filtering and noise reduction to determine the preprocessed image. Specifically, perform filtering and noise reduction processing on the transmitted image data to improve the quality of the image. This specifically includes using filtering algorithms such as median filtering and Gaussian filtering to remove noise and interference information in the image. Perform operations such as contrast enhancement and sharpening on the image to improve the clarity and distinguishability of the image.

[0040] Step S413: Traverse the risk database to match the preprocessed image and locate the hazard sources and hazard targets. Specifically, extract the template images of known hazard sources and hazard targets from the risk database and match them with the preprocessed image. Use the feature point matching algorithm to search for feature points similar to the template image in the image, so as to locate the hazard sources and hazard targets. Combine deep learning algorithms (such as convolutional neural network CNN) to identify and classify the targets in the image to further confirm the identities of the hazard sources and hazard targets. This implementation method improves the accuracy of identifying hazard sources and hazard targets in the image through the mining of historical risk control data and image preprocessing.

[0041] In a possible implementation manner, when performing the matching based on the two-dimensional projection space of the point cloud, step S420 further includes step S421: Based on the image space, frame the hazard source and determine the first coordinate, and frame the hazard target and determine the second coordinate. Specifically, frame the approximate areas of the hazard source and the hazard target in the image space. Extract the points of these areas as the first coordinate (hazard source coordinate) and the second coordinate (hazard target coordinate).

[0042] Step S422: According to the first coordinate and the second coordinate, perform coordinate positioning on the two-dimensional projection space of the point cloud to determine the first positioning coordinate. Specifically, use the projection plane to map the first coordinate and the second coordinate into the two-dimensional projection space of the point cloud to obtain the first positioning coordinate, that is, the first positioning coordinate includes the projection coordinates of the first coordinate and the second coordinate.

[0043] Step S423: Identify the pixel features of the hazard source and the hazard target, perform feature matching based on the feature descriptor, and determine the second matching coordinate. Specifically, extract the feature descriptors of the hazard source and the hazard target from the image. These feature descriptors can be local feature descriptors such as SIFT, SURF, ORB, etc., or global feature descriptors such as HOG, LBP, etc. Then, search for the point cloud features matching these feature descriptors in the two-dimensional projection space of the point cloud, and determine the matching result by calculating the similarity between the feature descriptors (such as Euclidean distance, Hamming distance, etc.). The matching result contains the coordinates of the point cloud features corresponding to the hazard source and the hazard target on the projection plane, that is, the second matching coordinate. The second matching coordinate also includes the matching coordinates of the first coordinate and the second coordinate.

[0044] Step S424, proofread and fuse the first positioning coordinate and the second matching coordinate to determine the projection matching result. Specifically, compare the differences between the first positioning coordinate and the second matching coordinate, and evaluate their accuracy and consistency. If the differences are large, it is necessary to re - perform feature matching or adjust the parameters of the coordinate mapping. Fuse the coordinates that are considered accurate and reliable after proofreading to obtain a final coordinate that combines the advantages of the two methods, and use this final coordinate as the basis for subsequent spatial distance measurement. This implementation method can accurately determine the positions of the hazard source and the dangerous target in the two - dimensional projection space of the point cloud by combining the two methods of coordinate positioning and feature matching. Coordinate positioning provides preliminary position information, and feature matching verifies and corrects this position information by comparing feature descriptors, thereby improving the accuracy of positioning.

[0045] In a possible implementation, for the spatial distance measurement, step S400 further includes step S440, identify the target point cloud positions, and determine the first point cloud range and the second point cloud range, where the first point cloud range is the point cloud distribution of the hazard source, and the second point cloud range is the point cloud distribution of the dangerous target. Specifically, the target point cloud positions are the specific positions of the hazard source and the dangerous target in the three - dimensional space of the point cloud. Determine their ranges by traversing these point clouds, specifically including calculating parameters such as the boundaries, centroids, and sizes of the point clouds. Among them, the point cloud range refers to the distribution area of the point clouds of the hazard source and the dangerous target in the three - dimensional space.

[0046] Step S450, based on the first point cloud range and the second point cloud range, determine the spatial distance measurement point cloud pairs, where the spatial distance measurement point clouds are boundary point cloud pairs. Specifically, by calculating methods such as the convex hull, minimum bounding box, and minimum enclosing sphere of the point cloud, extract the boundary point clouds from the point cloud ranges of the hazard source and the dangerous target (the outer point clouds of the point cloud ranges of the hazard source and the dangerous target, which define the boundaries of the target). Find the closest point pair between the two boundary point clouds as the spatial distance measurement point cloud pair by calculating the Euclidean distance between each pair of points.

[0047] Step S460, measure the relative distance of the spatial distance measurement point cloud pairs to determine the target spatial distance, where the relative distance includes the straight - line distance, horizontal distance, and vertical distance. Specifically, calculate the straight - line distance between the spatial distance measurement point cloud pairs by calculating the Euclidean distance between two points. Decompose the straight - line distance into horizontal and vertical components to calculate the horizontal distance and vertical distance. Finally, output the calculated relative distance as the target spatial distance for subsequent applications such as risk assessment and safety monitoring. This implementation method uses boundary point cloud pairs for distance measurement. The boundary point clouds are relatively stable and less affected by noise interference, improving the accuracy of measurement.

[0048] In a possible implementation, step S450 of determining the spatial ranging point cloud pair further includes step S451 of determining the central point cloud of the first point cloud range and the second point cloud range, and using the boundary point clouds on the line connecting the central point clouds as the first boundary point cloud and the second boundary point cloud. Specifically, by calculating the average coordinates of all points within the point cloud range, the central point cloud of the hazard source and the hazardous target point cloud range is calculated. The central point cloud represents the geometric center of the hazard source and the hazardous target. Then, based on the line connecting the central point clouds, the boundary point clouds on the line are found and used as the first boundary point cloud and the second boundary point cloud. These boundary point clouds are located at the edge of the hazard source and the hazardous target point cloud range and are the key points connecting the two target point clouds.

[0049] Step S452: Using the first boundary point cloud as a reference point, perform boundary extension measurement based on the reference point to determine the first effective point cloud of the hazard source, where the boundary geometric features are determined based on the point cloud normal vector and curvature. Specifically, using the first boundary point cloud as a reference point, utilize geometric features such as the normal vector and curvature of the point cloud to perform boundary extension measurement to determine the range of the effective point cloud. Within the boundary extension range, the effective point cloud is selected according to specific conditions (such as distance, angle, etc.) to determine the first effective point cloud of the hazard source. The effective point cloud is the point cloud obtained through boundary extension measurement and used for spatial distance measurement. Among them, the normal vector is a vector describing the surface direction of the point cloud, and the curvature is a quantity describing the degree of bending of the point cloud surface. By calculating the normal vector and curvature of the point cloud, the geometric features of the point cloud can be understood, thereby performing boundary extension measurement. Boundary extension measurement is an algorithm based on the geometric features of the point cloud used to determine the boundary of the point cloud range.

[0050] Step S453: Using the second boundary point cloud as a reference point, perform boundary extension measurement based on the reference point to determine the second effective point cloud of the hazardous target. Specifically, similarly using the second boundary point cloud as a reference point, perform boundary extension measurement to determine the second effective point cloud of the hazardous target.

[0051] Step S454: Use the first effective point cloud and the second effective point cloud as the boundary point cloud pair. Specifically, use the first effective point cloud and the second effective point cloud as the boundary point cloud pair for spatial distance measurement. The point cloud of the hazard source or the hazardous target may have an irregular distribution, so the boundary point cloud may not directly represent the shortest distance between the two targets. This implementation method can reduce the influence of these factors by determining the effective point cloud, more accurately find the shortest path between the two targets, and improve the accuracy and reliability of the calculation results.

[0052] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present application shall be included within the protection scope of the present application. In some cases, the actions or steps recited in the present application can be executed in an order different from that in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for measuring the distance between a hazard source and a target space using monocular vision, characterized in that, The method includes: Deploying front-end acquisition devices, where the front-end acquisition devices include a laser projection device and a vision camera; Controlling the laser projection device to perform a full-field scan of the target scene, transmitting back the laser point cloud data, controlling the vision camera to perform full-field acquisition of the scene, and transmitting back the image data, where the image data is a single image; Constructing a scene space model with the laser point cloud data and the image data, where the scene space model includes a point cloud three-dimensional space - a point cloud two-dimensional projection space - an image space; Through the scene space model, based on the matching of the image space - the point cloud two-dimensional projection space, and the point cloud mapping of the point cloud two-dimensional projection space - the point cloud three-dimensional space, determining the target point cloud position and performing spatial distance measurement to determine the target spatial distance, where the target point cloud position is the point cloud position of the hazard source and the dangerous target.

2. The method for measuring the distance between a hazard source and a target space using monocular vision according to claim 1, wherein, The constructing of the scene space model includes: Traversing the laser point cloud data to identify the point cloud coordinate data and the scene feature data of each laser point cloud; Based on the point cloud coordinate data, building the point cloud three-dimensional space, where the point cloud three-dimensional space represents the spatial position of the point cloud; Combining the scene feature data, using the point cloud three-dimensional space as a reference, constructing the point cloud two-dimensional projection space, where the point cloud two-dimensional projection space uses the field-of-view coordinate system of the image data as the projection plane; Building the image space with the image data, hierarchically integrating the point cloud three-dimensional space - the point cloud two-dimensional projection space - the image space, and determining the scene space model.

3. The method for measuring the spatial distance between a hazard source and a target using monocular vision according to claim 2, wherein, Combining the scene feature data, using the point cloud three-dimensional space as a reference, constructing the point cloud two-dimensional projection space includes: Determining the field-of-view coordinate system of the image data based on the acquisition parameters of the vision camera; According to the field-of-view coordinate system, constructing the projection plane, establishing the point cloud projection from the point cloud three-dimensional space to the projection plane, and determining the point cloud two-dimensional projection; Traversing the scene feature data, determining the feature descriptors of each laser point cloud, and performing projection point cloud identification on the point cloud two-dimensional projection to generate the point cloud two-dimensional projection space.

4. The method for measuring the distance between a hazard source and a target space using monocular vision according to claim 3, wherein The determining of the target point cloud position includes: Based on the image space, framing the hazard source and the dangerous target; For the hazard source and the dangerous target, performing matching based on the point cloud two-dimensional projection space to determine the projection matching result, which includes position matching and feature matching; Using the point cloud mapping of the point cloud two-dimensional projection space - the point cloud three-dimensional space to determine the three-dimensional mapped point cloud of the projection matching result as the target point cloud position.

5. The method for measuring the distance between a hazard source and a target space using monocular vision according to claim 4, characterized in that, The framing of the hazard source and the dangerous target includes: For the target scene, mining the risk database based on historical risk control data; Transmitting back the image data and performing filtering and noise reduction to determine the preprocessed image; Traversing the risk database, performing matching on the preprocessed image, and locating the hazard source and the dangerous target.

6. The method for measuring the distance between a hazard source and a target space using monocular vision according to claim 4, wherein Performing matching based on the point cloud two-dimensional projection space includes: Based on the image space, framing the hazard source and determining the first coordinate, framing the dangerous target and determining the second coordinate; Based on the first coordinate and the second coordinate, perform coordinate positioning on the two-dimensional projection space of the point cloud to determine the first positioning coordinate; Identify the pixel features of the hazard source and the hazard target, perform feature matching based on the feature descriptor, and determine the second matching coordinate; Proofread and fuse the first positioning coordinate and the second matching coordinate to determine the projection matching result.

7. The method for measuring the distance between a hazard source and a target space using monocular vision according to claim 1, characterized in that, The performing of the spatial distance measurement includes: Identify the position of the target point cloud, determine the first point cloud range and the second point cloud range, where the first point cloud range is the point cloud distribution of the hazard source, and the second point cloud range is the point cloud distribution of the hazard target; Based on the first point cloud range and the second point cloud range, determine the spatial distance measurement point cloud pair, where the spatial distance measurement point cloud is the boundary point cloud pair; Measure the relative distance of the spatial distance measurement point cloud pair to determine the target spatial distance, where the relative distance includes the straight-line distance, the horizontal distance, and the vertical distance.

8. The method for measuring the distance between a hazard source and a target space using monocular vision according to claim 7, characterized in that, The determining of the spatial distance measurement point cloud pair includes: Determine the center point cloud of the first point cloud range and the second point cloud range, and use the boundary point cloud on the center point cloud connection line as the first boundary point cloud and the second boundary point cloud; Taking the first boundary point cloud as the reference point, perform boundary extension measurement based on the reference point to determine the first effective point cloud of the hazard source, where the boundary geometric features are determined based on the point cloud normal vector and curvature; Taking the second boundary point cloud as the reference point, perform boundary extension measurement based on the reference point to determine the second effective point cloud of the hazard target; Use the first effective point cloud and the second effective point cloud as the boundary point cloud pair.

Images