Vectorization method and device for point cloud data

Abstract

The embodiment of the invention discloses a point cloud data vectorization method and device, and the method comprises the steps: carrying out the classification of a plane structure of a target indoor space based on a point cloud data image of the target indoor space, and obtaining a plurality of types of plane structures; determining a target plane structure from the plurality of types of plane structures; and performing graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space so as to express a topological structure of the target indoor space through the vectorized data structure. Therefore, a full-automatic vectorization scheme for the point cloud data image of the target indoor space is realized, the topographic line drawing efficiency of the target indoor space can be improved, and the consistency of the measurement size is ensured, so that the topographic line drawing accuracy of the target indoor space is improved.

Description

Technical Field

[0001] The present invention relates to three-dimensional reconstruction technology and point cloud processing technology, and in particular to a method and device for vectorizing point cloud data. Background Art

[0002] The point cloud data of indoor scenes is a collection of three-dimensional coordinate points, representing the objects and structures in the indoor scenes. The vectorization of point cloud data of indoor scenes is to convert these point cloud data into vector data for better spatial analysis and processing.

[0003] In the related art, the means of vectorizing point cloud data of indoor scenes mostly require first cropping the point cloud data to ensure that the walls, doors, doorways and other structures shown in the point cloud data can be clearly identified from a bird's-eye view. Then, based on the image from a bird's-eye view, lines are manually drawn along these structures, and finally connected into several topologically connected line segments to complete the vectorization work. This method has at least the following problems: the manual mapping and tracing of structural lines is usually inefficient; it relies heavily on the habits and experience of the operator, and the accuracy variance of tracing structural lines may be relatively large. Summary of the invention

[0004] In order to solve the technical problems in the related art, the embodiments of the present disclosure provide a method and device for vectorizing point cloud data.

[0005] According to a first aspect of an embodiment of the present disclosure, a method for vectorizing point cloud data is provided, the method comprising:

[0006] Based on the point cloud data image of the target indoor space, the plane structure of the target indoor space is classified to obtain plane structures of multiple categories;

[0007] Determine a target plane structure from the plane structures of the plurality of categories;

[0008] Determine a top view image of the point cloud data image of the target indoor space;

[0009] In the overhead image, performing structural completion processing on all plane structures of the target indoor space to obtain the completed plane structure of the target indoor space;

[0010] The target plane structure is processed by a graph cut algorithm to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space, so as to express the topological structure of the target indoor space through the vectorized data structure.

[0011] As an optional embodiment of the present disclosure, determining the target plane structure from the multiple categories of plane structures includes:

[0012] All plane structures of the multiple categories are labeled, wherein the plane structures labeled with the first category label are target plane structures.

[0013] As an optional embodiment of the present disclosure, the step of performing graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space includes:

[0014] Based on the plane structure of the target indoor space after the completion, determining a solution space formed by the target plane structure, wherein the solution space is the image to be segmented processed by the graph cut algorithm, and is used to solve the cut points and cut edges processed by the graph cut algorithm;

[0015] In the solution space, visibility information of the point cloud data of the target indoor space is used to determine a vectorized data structure corresponding to the point cloud data image of the target indoor space.

[0016] As an optional embodiment of the present disclosure, determining a solution space formed by the target plane structure based on the completed plane structure of the target indoor space includes:

[0017] Determine a first outer rectangular contour according to an outer bounding box of the point cloud data of the target indoor space;

[0018] Determining a target plane structure from the completed plane structure of the target indoor space;

[0019] A solution space formed by the target plane structure is determined based on the first outer rectangular outline and the target plane structure.

[0020] As an optional embodiment of the present disclosure, determining a solution space formed by the target plane structure based on the first outer rectangular contour and the target plane structure includes:

[0021] Enlarging and adjusting the first outer rectangular outline according to a preset size to obtain a second outer rectangular outline;

[0022] Determining a solution space formed by the target plane structure based on the second outer rectangular contour and the target plane structure;

[0023] Wherein, the first outer rectangular outline and the second outer rectangular outline are both rectangular polygons.

[0024] As an optional embodiment of the present disclosure, determining a solution space formed by the target plane structure based on the second outer rectangular contour and the target plane structure includes:

[0025] Determine a top-down projection diagram of the target plane structure, and obtain a two-dimensional line segment corresponding to each of the target plane structures;

[0026] Inserting the two-dimensional line segment corresponding to each of the target plane structures into the rectangular polygon formed by the second outer rectangular contour to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygon areas, wherein the plurality of rectangular polygon areas constitute the solution space.

[0027] As an optional embodiment of the present disclosure, the step of inserting the two-dimensional line segment corresponding to each of the target plane structures into the rectangular polygon formed by the second outer rectangular contour to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygon areas includes:

[0028] Determine the length value of the two-dimensional line segment corresponding to each of the target plane structures;

[0029] Based on the length values ​​of the two-dimensional line segments corresponding to all the target plane structures, the two-dimensional line segments are inserted into the rectangular polygon formed by the second outer rectangular outline in order from long to short, so as to divide the rectangular polygon formed by the second outer rectangular outline into multiple rectangular polygon areas.

[0030] As an optional embodiment of the present disclosure, in the solution space, using visibility information of the point cloud data of the target indoor space to determine a vectorized data structure corresponding to the point cloud data image of the target indoor space includes:

[0031] Determining the coordinate position of an image acquisition device in the solution space, wherein the image acquisition device is used to acquire point cloud data of the target indoor space;

[0032] Connecting the coordinate position of the image acquisition device to any point of the point cloud data of the target indoor space to obtain a plurality of connecting line segments;

[0033] Setting weights for a plurality of rectangular polygonal areas in the solution space through which each of the connecting line segments passes according to a preset weight setting method;

[0034] Dividing a plurality of rectangular polygonal areas in the solution space into indoor space and outdoor space according to weight settings of the rectangular polygonal areas in the solution space;

[0035] The rectangular polygonal areas of the indoor space and the outdoor space are sequentially connected to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space.

[0036] As an optional embodiment of the present disclosure, the step of setting weights for the plurality of rectangular polygonal areas in the solution space through which each connecting line segment passes according to a preset weight setting method includes:

[0037] Determine the vertex of the graph corresponding to the rectangular polygonal area where the image acquisition device is located as the first vertex;

[0038] Determine an edge obtained by connecting the first vertex to the starting point determined by the graph cut algorithm as a first target edge;

[0039] Adding a fixed weight value to an original weight value of the first target edge to obtain a final weight value of the first target edge;

[0040] For any one of the plurality of connecting line segments, determining a plurality of common edges involved in a rectangular polygonal area passed through by the connecting line segment as second target edges;

[0041] Adding the fixed weight value to the original weight value of the second target edge to obtain a final weight value of the second target edge;

[0042] Determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which the point cloud data of the target indoor space in the solution space belongs;

[0043] Determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which any point cloud data belongs in the solution space as the third target edge;

[0044] The fixed weight value is subtracted from the original weight value of the third target edge to obtain a final weight value of the third target edge.

[0045] According to a second aspect of an embodiment of the present disclosure, a device for vectorizing point cloud data is provided, the device comprising:

[0046] A plane structure classification module is used to classify the plane structure of the target indoor space based on the point cloud data image of the target indoor space to obtain plane structures of multiple categories;

[0047] A first determining module, configured to determine a target plane structure from the plane structures of the plurality of categories;

[0048] A second determination module is used to determine a top view image of the point cloud data image of the target indoor space before performing the graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space;

[0049] A structure completion module, used for performing structure completion processing on all plane structures of the target indoor space in the top view image to obtain the completed plane structure of the target indoor space;

[0050] The graph cut processing module is used to perform graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space, so as to express the topological structure of the target indoor space through the vectorized data structure.

[0051] As an optional embodiment of the present disclosure, the first determining module includes:

[0052] The labeling unit is used to label all the plane structures of the multiple categories, wherein the plane structures labeled with the first category label are target plane structures.

[0053] As an optional embodiment of the present disclosure, the graph cut processing module includes:

[0054] A solution space determination unit, used to determine a solution space formed by the target plane structure based on the plane structure of the target indoor space after the completion, wherein the solution space is the image to be segmented processed by the graph cut algorithm, and is used to solve the cut points and cut edges processed by the graph cut algorithm;

[0055] The vectorized data determination unit is used to determine the vectorized data structure corresponding to the point cloud data image of the target indoor space in the solution space by using the visibility information of the point cloud data of the target indoor space.

[0056] As an optional embodiment of the present disclosure, the solution space determining unit includes:

[0057] A first determining subunit, configured to determine a first outer rectangular contour according to an outer bounding box of the point cloud data of the target indoor space;

[0058] A second determining subunit is used to determine a target plane structure from the completed plane structure of the target indoor space;

[0059] The third determining subunit is configured to determine a solution space formed by the target plane structure based on the first outer rectangular contour and the target plane structure.

[0060] As an optional embodiment of the present disclosure, the third determining subunit is further configured to:

[0061] Enlarging and adjusting the first outer rectangular outline according to a preset size to obtain a second outer rectangular outline;

[0062] Determining a solution space formed by the target plane structure based on the second outer rectangular contour and the target plane structure;

[0063] Wherein, the first outer rectangular outline and the second outer rectangular outline are both rectangular polygons.

[0064] As an optional embodiment of the present disclosure, the third determining subunit implements the step of determining a solution space formed by the target plane structure based on the second outer rectangular contour and the target plane structure, including:

[0065] Determine a top-down projection diagram of the target plane structure, and obtain a two-dimensional line segment corresponding to each of the target plane structures;

[0066] Inserting the two-dimensional line segment corresponding to each of the target plane structures into the rectangular polygon formed by the second outer rectangular contour to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygon areas, wherein the plurality of rectangular polygon areas constitute the solution space.

[0067] As an optional embodiment of the present disclosure, the third determining subunit implements the inserting of the two-dimensional line segment corresponding to each of the target plane structures into the rectangular polygon formed by the second outer rectangular outline to divide the rectangular polygon formed by the second outer rectangular outline into a plurality of rectangular polygon areas, including:

[0068] Determine the length value of the two-dimensional line segment corresponding to each of the target plane structures;

[0069] Based on the length values ​​of the two-dimensional line segments corresponding to all the target plane structures, the two-dimensional line segments are inserted into the rectangular polygon formed by the second outer rectangular outline in order from long to short, so as to divide the rectangular polygon formed by the second outer rectangular outline into multiple rectangular polygon areas.

[0070] As an optional embodiment of the present disclosure, the vectorized data determining unit includes:

[0071] a fourth determination subunit, configured to determine a coordinate position of an image acquisition device in the solution space, wherein the image acquisition device is configured to acquire point cloud data of the target indoor space;

[0072] A connecting subunit, used to connect the coordinate position of the image acquisition device to any point of the point cloud data of the target indoor space to obtain a plurality of connecting line segments;

[0073] A weight setting subunit, used for setting weights for a plurality of rectangular polygonal areas in the solution space passed by each of the connecting line segments according to a preset weight setting method;

[0074] A region division subunit, configured to divide a plurality of rectangular polygonal regions in the solution space into indoor space and outdoor space according to weight settings of the rectangular polygonal regions in the solution space;

[0075] The vectorized data determination subunit is used to sequentially connect the rectangular polygonal areas of the indoor space and the outdoor space to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space.

[0076] As an optional embodiment of the present disclosure, the weight setting subunit is further used to:

[0077] Determine the vertex of the graph corresponding to the rectangular polygonal area where the image acquisition device is located as the first vertex;

[0078] Determine an edge obtained by connecting the first vertex to the starting point determined by the graph cut algorithm as a first target edge;

[0079] Adding a fixed weight value to an original weight value of the first target edge to obtain a final weight value of the first target edge;

[0080] For any one of the plurality of connecting line segments, determining a plurality of common edges involved in a rectangular polygonal area passed through by the connecting line segment as second target edges;

[0081] Adding the fixed weight value to the original weight value of the second target edge to obtain a final weight value of the second target edge;

[0082] Determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which the point cloud data of the target indoor space in the solution space belongs;

[0083] Determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which any point cloud data belongs in the solution space as the third target edge;

[0084] The fixed weight value is subtracted from the original weight value of the third target edge to obtain a final weight value of the third target edge.

[0085] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0086] a memory for storing a computer program product;

[0087] The processor is used to execute the computer program product stored in the memory, and when the computer program product is executed, the method described in the first aspect above is implemented.

[0088] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described in the first aspect above is implemented.

[0089] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, including computer program instructions, which, when executed by a processor, implement the method described in the first aspect above.

[0090] Through the technical solution of the disclosed embodiment, based on the point cloud data image of the target indoor space, the plane structure of the target indoor space is classified, and then the target plane structure is determined from multiple categories of plane structures, and then the target plane structure is processed by the graph cutting algorithm to obtain the vectorized data structure corresponding to the point cloud data image of the target indoor space. Thus, a fully automated vectorization solution for the point cloud data image of the target indoor space is realized, which can improve the efficiency of mapping and tracing the target indoor space, while ensuring the consistency of the measured dimensions, avoiding the measurement accuracy variance caused by the operator's manual measurement, thereby improving the accuracy of mapping and tracing the target indoor space. In addition, by performing structural completion processing on all the plane structures of the target indoor space, a more complete plane structure can be obtained, further improving the accuracy of the vectorization of the point cloud data.

[0091] The technical solution of the present disclosure is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0093] The present disclosure may be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

[0094] Figure 1 The present invention is a flowchart of a method for vectorizing point cloud data according to an embodiment of the method disclosed herein.

[0095] Figure 2 The present invention is a flowchart of a method for vectorizing point cloud data according to another embodiment of the present invention.

[0096] Figure 3 A schematic diagram of the effect of plane structure padding processing in one embodiment of the method disclosed herein.

[0097] Figure 4 A schematic diagram of the effect of determining the solution space of an embodiment of the method disclosed herein.

[0098] Figure 5 A schematic diagram of weight settings for an embodiment of the method disclosed herein.

[0099] Figure 6 A schematic diagram of a vectorized data structure of an indoor space according to an embodiment of the method disclosed herein.

[0100] Figure 7 The present invention is a structural block diagram of a point cloud data vectorization device according to an embodiment of the present invention.

[0101] Figure 8 The present invention is a block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0102] In the embodiments of the present disclosure, an image acquisition device such as a laser radar can be used to perform an all-round scan of the indoor scene or indoor space to obtain point cloud data of surfaces such as walls, ceilings, and floors, thereby obtaining a point cloud data image of the indoor space. It can be understood that the point cloud data image of the indoor space is composed of a large number of discrete points. There is no topological relationship between these discrete points, but each point contains spatial coordinates and possible other attribute information. Therefore, the embodiments of the present disclosure use the spatial coordinates of the point cloud data of the indoor space, combined with the graph cut algorithm processing technology, to establish the topological relationship of the point cloud data of the indoor space, obtain the tracing line of the indoor space, and realize the vectorization of the point cloud data of the indoor space.

[0103] Among them, the graph cut algorithm processing technology is an image segmentation method based on graph theory, and its core idea is to use the minimum cut model in graph theory to find the optimal segmentation in the image. The graph cut algorithm constructs a graph in which nodes represent pixels or regions in the image, and edges represent the similarities or differences between pixels, and then segments the image by finding the edge weights and the minimum cut. In the embodiment of the present disclosure, a two-dimensional overhead image formed based on point cloud data of an indoor space is segmented by a graph cut algorithm to divide the two-dimensional overhead image into indoor space and outdoor space, thereby obtaining a tracing line of the indoor space and realizing vectorization of the point cloud data.

[0104] In order for those skilled in the art to clearly understand the technical solution of the embodiments of the present disclosure, the technical solution of the embodiments of the present disclosure is described in detail below with reference to the accompanying drawings.

[0105] Figure 1 FIG. 1 is a flow chart of a method for vectorizing point cloud data according to an embodiment of the method disclosed herein. Figure 1 As shown, a method for vectorizing point cloud data in an embodiment of the present disclosure may include the following steps:

[0106] Step 101 , based on the point cloud data image of the target indoor space, the plane structure of the target indoor space is classified to obtain plane structures of multiple categories.

[0107] The target indoor space is an indoor environment whose point cloud data image is to be vectorized, for example, a room. In the disclosed embodiment, an image acquisition device such as a laser radar can be used to perform an all-round scan of each plane of the target indoor space, such as the ceiling, floor, wall, etc. of the target indoor space, to obtain a point cloud data image of the target indoor space. It is understandable that in addition to its own structure (ceiling, floor, wall), the target indoor space may also include ancillary structures such as door panels, wardrobes, cabinets, refrigerators, etc. Therefore, the point cloud data will also include the point cloud data of these ancillary structures.

[0108] It can be understood that the structure of the indoor space is mostly composed of planes. In this step, the point cloud data of the indoor space can be first divided into several plane structures. Each plane structure has corresponding point cloud data and plane structure parameters. The plane structure parameters include but are not limited to the normal vector and the center point position. Then, based on the relationship between the normal vector of the plane structure and the gravity direction, the plane structure of the target indoor space can be classified to obtain multiple categories of plane structures.

[0109] Specifically, the plane structures of the target indoor space are divided into four categories: wall-like, ceiling-like, floor-like, and unknown. For example, for any plane structure, if its normal vector direction is perpendicular to the gravity direction, it is classified as a wall-like structure (such as a wall, door panel, wardrobe, etc.), if its normal vector direction is consistent with the gravity direction, it is classified as a ceiling-like structure, if its normal vector direction is opposite to the gravity direction, it is classified as a floor-like structure (such as the ground, bed surface, desktop, etc.), and other plane structures are classified as unknown types.

[0110] Step 102: determine a target plane structure from multiple categories of plane structures.

[0111] In the embodiment of the present disclosure, priority is given to two-dimensional vectorization of the structural parts of the target indoor space itself (such as walls, doors, doorways), while its auxiliary structural parts (such as wardrobes, refrigerators, etc.) do not need to be vectorized. Therefore, it is necessary to filter some structures that do not need to be vectorized. As an optional embodiment of the present disclosure, it includes but is not limited to the following filtering conditions: only plane structures of the type of wall-like and the structural parts of the target indoor space itself are taken, and the structural parts of the non-target indoor space such as door panels, wardrobes, cabinets, refrigerators, etc. are filtered out to prevent them from affecting the accuracy of the structural parts themselves. That is, the plane structure of the type of wall-like and the structural part of the target indoor space itself is determined as the target plane structure.

[0112] As an implementation method of the embodiment of the present disclosure, all plane structures of multiple categories can be labeled, wherein the plane structure labeled with the first type of label is the target plane structure. For example, a plane structure that meets the target plane structure condition is labeled as "1", and a plane structure that does not meet the target plane structure condition is labeled as "0".

[0113] Specifically, in the disclosed embodiments, the size and position of the plane structure can be used, for example, on a top view of the point cloud data image of the target indoor space, according to the point cloud position and size of any plane structure, to distinguish whether it is a plane structure that meets the target plane structure conditions. For example, if the point cloud position of the wall is at the edge of the point cloud image, and its size, such as height, is the largest, it is marked as, for example, "1" (i.e., it meets the target plane structure conditions), and for example, for a wardrobe that does not reach the ceiling, its point cloud position is actually at an internal position, and its size is smaller than the wall size, it is marked as, for example, "0" (i.e., it does not meet the target plane structure conditions). In order to completely retain the point cloud data, the target plane structure is determined by marking instead of directly deleting the point cloud data of the non-target plane structure.

[0114] Step 103: determine a bird's-eye view image of the point cloud data image of the target indoor space.

[0115] The point cloud data image of the target indoor space can be projected from a bird's-eye view to obtain a bird's-eye view image of the point cloud data image of the target indoor space. In the bird's-eye view image, the point cloud data images of each plane structure in the target indoor space will form a number of corresponding line segments to form a two-dimensional polygonal image.

[0116] Step 104: In the bird's-eye view image, all plane structures of the target indoor space are structure-completed to obtain the completed plane structure of the target indoor space.

[0117] Since image acquisition devices such as LiDAR often have problems such as incomplete scanning and occlusion in actual point cloud scanning, this will actually cause some data of extracted plane structures to be missing in the point cloud data. The line segments formed by the plane structure may not intersect with the originally intersecting line segments, making the two-dimensional polygonal image not a closed image or a missing two-dimensional image. Therefore, in order to ensure that the plane structure extracted from the point cloud data can be connected into a closed polygon, that is, a closed two-dimensional polygonal image, the plane structure needs to be filled.

[0118] As an implementable method of the disclosed embodiment, in a top-view image, a first supplementary plane is added that is perpendicular to both the normal vector direction and the gravity direction of any planar structure and passes through any endpoint of a top-view line segment of any planar structure, wherein the first supplementary plane is used to structurally supplement any planar structure. This embodiment is based on the Manhattan hypothesis (whose basic assumption is that objects are composed of cuboids and parallelograms), and according to the existing planar structure of the target indoor space, for each planar structure, two planes are added that are perpendicular to both the normal vector direction and the gravity direction of the planar structure and pass through any of the two endpoints of the corresponding line segment in the top-view image of the planar structure as supplements. For details, see Figure 3 In the figure, the gray line segments are the top-view segments corresponding to each plane structure, and the black line segments are the top-view segments corresponding to the corresponding supplementary planes. Figure 3 The top-down line segment of the plane structure numbered 1 and the black line segment perpendicular to it at its endpoint A are one of the supplementary planes that are filled in based on its normal vector, the direction of re-erection and the endpoint A.

[0119] As another implementable method of the embodiment of the present disclosure, the 2.5D hypothesis (i.e., the three-dimensional structure in the room can be described by giving height information using the two-dimensional image on the overhead image) can be used to extract the line segments on the outline of the overhead image through the overhead image of the point cloud data image, and all the extracted line segments are stretched into three-dimensional space to make them supplementary planes of the corresponding plane structures. Specifically, in the overhead image, for the contour line segments of any planar structure in the overhead image, according to the height value in the direction of the preset target coordinate axis (such as the z-axis), the contour line segments of any planar structure can be stretched by a stretching tool (such as the stretching operation in Opencv) to obtain the second supplementary plane corresponding to the contour line segments of any planar structure, wherein the second supplementary plane is used to structurally supplement the planar structure corresponding to the contour line segments of any planar structure.

[0120] In the embodiment of the present disclosure, only the target plane structure can be plane-complemented, or all the plane structures of the target indoor space can be plane-complemented. Due to the complete lack of some plane structures, their supplementation may rely on two plane structures. For example, the wall structure between the wardrobe and the door depends on the two plane structures of the wardrobe and the door panel. Therefore, if only the target plane structure is supplemented, the wall structure may not be supplemented. In order to ensure that all the plane structures of the target indoor space can be restored and improve the accuracy of vectorization, the embodiment of the present disclosure takes plane-complementing all the plane structures of the target indoor space as the preferred implementation scheme.

[0121] As another feasible embodiment, a plane supplement scheme based on the Manhattan hypothesis and a plane supplement scheme based on the 2.5D hypothesis can be combined. The former performs plane supplementation through the extracted plane, and the latter performs plane supplementation through the extracted top view contour line. The two complement each other, making the supplemented plane structure more complete, thereby providing a guarantee for the accuracy of vectorization.

[0122] Step 105 , performing graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space, so as to express the topological structure of the target indoor space through the vectorized data structure.

[0123] In the disclosed embodiment, these finite two-dimensional planes of the target plane structure can be a two-dimensional space. If the two-dimensional space has n target plane structures, the two-dimensional space is divided into two inner and outer spaces by the target plane structure. When n target plane structures are inserted into the two-dimensional space, it is divided into several polygons. Each polygon corresponds to a vertex of the graph cut algorithm. Graph cut edges are established between polygons with common edges. Then, a polygonal area is solved through effective weight setting. The interior of the polygonal area is the indoor space, and the outside of the polygonal area is the external space. The boundary between the inner and outer spaces is a closed polygon, which completely corresponds to the structure of the point cloud, thereby completing the vectorization of the point cloud data.

[0124] In summary, based on the point cloud data image of the target indoor space, the plane structure of the target indoor space is classified, and then the target plane structure is determined from multiple categories of plane structures, and then the target plane structure is processed by the graph cutting algorithm to obtain the vectorized data structure corresponding to the point cloud data image of the target indoor space, thereby realizing a fully automated vectorization solution for the point cloud data image of the target indoor space, improving the efficiency of mapping and tracing the target indoor space, and ensuring the consistency of the measured dimensions, avoiding the measurement accuracy variance caused by manual measurement by the operator, and improving the accuracy of mapping and tracing the target indoor space. In addition, by performing structural completion processing on all plane structures of the target indoor space, a more complete plane structure can be obtained, further improving the accuracy of vectorization of point cloud data.

[0125] As another possible implementation method of the present disclosure, Figure 2 FIG. 1 is a flow chart of a method for vectorizing point cloud data according to another embodiment of the method disclosed herein. Figure 2 As shown, in Figure 1 Based on the illustrated embodiment, the aforementioned step 105 can be implemented as follows:

[0126] Step 1051, based on the completed plane structure of the target indoor space, determine the solution space formed by the target plane structure, wherein the solution space is the image to be segmented processed by the graph cut algorithm, and is used to solve the cut points and cut edges processed by the graph cut algorithm.

[0127] According to an achievable embodiment of the present disclosure, a first outer rectangular contour can be determined based on the outer bounding box of the point cloud data of the target indoor space, a target plane structure can be determined from the completed plane structure of the target indoor space, and then a solution space formed by the target plane structure can be determined based on the first outer rectangular contour and the target plane structure.

[0128] Among them, the outer bounding box of the point cloud is a geometric shape used to approximate the boundary range of the point cloud data, and is mainly used for fast analysis and processing of point cloud data. Common point cloud outer bounding box types include Axis-Aligned Bounding Box (AABB) and Oriented Bounding Box (OBB). AABB (Axis-Aligned Bounding Box) is an axis-aligned bounding box whose boundaries are parallel to the coordinate axes. The calculation of AABB is relatively simple, but it may not fit the shape of the point cloud closely. OBB (Oriented Bounding Box) can be rotated according to the direction of the point cloud to fit the shape of the point cloud more closely. The calculation of OBB is relatively complex, but it can provide more accurate boundary information. The disclosed embodiment uses the outer bounding box of point cloud data to quickly determine the boundary range of point cloud data and improve processing efficiency.

[0129] According to an achievable embodiment of the present disclosure, on the basis of the first outer rectangular contour, the first outer rectangular contour can be enlarged and adjusted according to a preset size to obtain a second outer rectangular contour, and then the solution space formed by the target plane structure is determined based on the second outer rectangular contour and the target plane structure, wherein the first outer rectangular contour and the second outer rectangular contour are both rectangular polygons. That is, the outermost rectangular contour (rectangular polygon) of the solution space is determined according to the outer bounding box of the point cloud data of the target indoor space, and the outermost rectangular contour is further expanded to a certain extent compared to the compact bounding box (that is, closely attached to the boundary of the point cloud data), so as to reserve space for the description of the outdoor space. Exemplarily, the preset size is, for example, 5 to 10 mm.

[0130] Furthermore, determining the solution space constituted by the target plane structure based on the second outer rectangular contour and the target plane structure can be achieved through the following steps: first, the top-down projection image of the target plane structure can be determined to obtain the two-dimensional line segments corresponding to each target plane structure, and then the two-dimensional line segments corresponding to each target plane structure are inserted into the rectangular polygon constituted by the second outer rectangular contour to divide the rectangular polygon constituted by the second outer rectangular contour into multiple rectangular polygonal areas, wherein the multiple rectangular polygonal areas constitute the solution space.

[0131] As an optional implementation method of the present embodiment, the two-dimensional line segments corresponding to each target plane structure are inserted into the rectangular polygon formed by the second outer rectangular contour. The length values ​​of the two-dimensional line segments corresponding to each target plane structure can be determined first, and based on the length values ​​of the two-dimensional line segments corresponding to all target plane structures, the two-dimensional line segments are inserted into the rectangular polygon formed by the second outer rectangular contour in order from long to short, so as to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygonal areas. It can be understood that, for example, if the target plane structure is a wall-like plane area or plane structure, in the embodiment of the present disclosure, the projection of the wall-like plane area on the horizontal plane can be formed into a two-dimensional line segment, and then inserted into the rectangular polygon in order from long to short. One line segment will cut the intersecting polygons into two polygons, and so on. The outermost rectangular polygonal area will be cut into a plurality of polygonal areas. These polygonal areas have common edges, and the common edges all come from the projection of the inserted wall-like plane area on the horizontal plane into a two-dimensional line segment. See, for example, Figure 4 As shown, the rectangular polygon formed by the second outer rectangular outline is Figure 4 The outermost and largest rectangular frame A is shown in the figure, and the inserted two-dimensional line segments are the line segments numbered ① to ⑤. These inserted line segments cut the rectangular frame A (rectangular polygon) into multiple polygonal areas.

[0132] The multiple polygonal areas obtained are the solution space, that is, the graph in the graph cut algorithm. Those skilled in the art can understand that a graph cut is composed of vertices and edges connecting vertices, such as n vertices and k edges. The graph cut algorithm divides the graph into two categories. In the embodiment disclosed in the present invention, the indoor space is divided into internal space and external space, and then a polygon is judged to belong to the internal space or the external space. Correspondingly, each polygon is used as a vertex of the graph of the graph cut algorithm. When polygons are adjacent to each other, there will be a common edge, which is the edge of the graph cut. In this way, the solution space formed by the polygonal area is mapped to the graph cut algorithm, and the polygonal area and the vertices and edges in the graph cut algorithm are determined.

[0133] Step 1052: In the solution space, the visibility information of the point cloud data of the target indoor space is used to determine the vectorized data structure corresponding to the point cloud data image of the target indoor space.

[0134] In the disclosed embodiment, the visibility information of the point cloud is used as the basis for calculation, that is, the posture information of the image acquisition device associated with the point cloud when it is collected. It can be understood that in this indoor space, there is no entity where the line segment connecting each point cloud in the point cloud data and the image acquisition device associated with it passes. Specifically, the point cloud visibility information means that when shooting, if an image acquisition device such as a lidar can collect any point, it can be determined that the line between the position of the image acquisition device and the point is definitely unobstructed. Each collected point cloud will have a camera position. Using this information, in the top view, the polygonal area passed by the line between the position of the image acquisition device and the point is determined as the indoor space, and the polygonal area not passed is determined as the external space.

[0135] Exemplarily, as an implementable solution of the embodiment of the present disclosure, first, the coordinate position of the image acquisition device in the solution space can be determined, and the image acquisition device is used to acquire the point cloud data of the target indoor space. The coordinate position of the image acquisition device is connected to any point of the point cloud data of the target indoor space, so that a plurality of connecting line segments can be obtained. The weights of the plurality of rectangular polygonal areas in the solution space passed by each connecting line segment are set according to a preset weight setting method, and then according to the weight setting of the rectangular polygonal areas in the solution space, the plurality of rectangular polygonal areas in the solution space are divided into indoor space and outdoor space, and the rectangular polygonal areas of the indoor space and the outdoor space are connected in turn to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space.

[0136] Furthermore, in the embodiment of the present disclosure, weights are set for multiple rectangular polygonal areas in the solution space passed by each connecting line segment according to a preset weight setting method. First, the vertex of the graph corresponding to the rectangular polygonal area where the image acquisition device is located can be determined as the first vertex, and the edge obtained by connecting the first vertex to the starting point determined by the graph cut algorithm is determined as the first target edge. A fixed weight value is added to the original weight value of the first target edge to obtain the final weight value of the first target edge; then, for any one of the multiple connecting line segments, multiple common edges involved in the rectangular polygonal area passed by the connecting line segment are determined as the second target edge, and a fixed weight value is added to the original weight value of the second target edge to obtain the final weight value of the second target edge; finally, the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which the point cloud data of the target indoor space in the solution space belongs is determined, and the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which any point cloud data in the solution space belongs is determined as the third target edge, and the fixed weight value is subtracted from the original weight value of the third target edge to obtain the final weight value of the third target edge.

[0137] For example, the image acquisition device (such as Figure 5 Point A in the figure must be located inside the indoor space, so the vertices of the graph corresponding to the rectangular polygonal area where the image acquisition device is located (such as the rectangular polygonal area numbered ① in the figure), and the vertex connection graph source (i.e., the starting point, such as Figure 5 The original weight (e.g., inf) of S in the image acquisition device is increased by a fixed weight (e.g., w), that is, the final weight value of the first target edge is obtained, wherein the original weight inf is a value much larger than the fixed weight w; the image acquisition device is moved to any point cloud (e.g., Figure 5 The points a, b, c, d, e, f, g, h, i, and j in the figure are connected into a line segment. The weight of the edge of the corresponding graph of the common edge passed by this line segment will increase by a fixed weight (for example, w), which is the final weight value of the second target edge, such as Figure 5 The rectangular polygonal area that point A passes through to connect point i includes the rectangular polygonal areas numbered 1, 2, 8, 9, 10, and 11. The common edges between these rectangular polygonal areas are the edges of the corresponding graph, and their corresponding weight values ​​will increase the fixed weight (such as w); while the point cloud (such as Figure 5 The edge of the graph to which the points a, b, c, d, e, f, g, h, i, j belong) will be given a reverse fixed weight (e.g., w') from outside to inside, that is, the final weight value of the third target edge is obtained. The reverse fixed weight (e.g., w') represents the part with the smallest weight between the starting point S and the end point T in the graph cut algorithm, that is, the position that is easy to disconnect. In the present disclosure, Figure 5 The common edges of the rectangular polygonal areas numbered 11 and 12, and the common edges of the rectangular polygonal areas numbered 5 and 6, occupied by the midpoint cloud.

[0138] In this way, by setting the weights in the solution space, multiple rectangular polygonal areas are finally divided into internal space and external space, and then the edges of the rectangular polygonal areas between the inside and outside are connected in sequence to form a closed polygonal outline, which is the vectorized data structure of the point cloud data of the target indoor space in the embodiment of the present disclosure. In the embodiment of the present disclosure, the vectorized data structure can be expressed by points, lines and areas, for example:

[0139] Point: the vertex of all contour polygons;

[0140] Line (e.g. wall line): It consists of the vertex indices of the two endpoints. The index corresponds to the serial number of the point in the set of points. If the index is unique, then all wall lines can be uniquely determined. The index of the wall line is the default serial number formed in the process of connecting the serial numbers of the points into a line.

[0141] Area (internal space, external space): It is composed of a set of wall line indexes. A closed polygon can be obtained by connecting the wall lines in sequence according to the index order.

[0142] In the disclosed embodiment, since each line segment includes two endpoints, as a preferred solution, only the endpoint coordinate values ​​of the line segments in each area can be stored, and each endpoint coordinate value can be marked with a serial number, that is, each list of vertices of the contour polygon has a corresponding serial number, so that when describing the line segment, it can be a vertex number to a vertex number, and the vertices with corresponding serial numbers can be connected to obtain the line segment (i.e., the wall line). In this way, only the endpoint coordinate values ​​and the corresponding needs are stored, which can reduce the consumption of storage resources.

[0143] Furthermore, the corresponding vertices to be connected are connected to obtain line segments (i.e. wall lines), and then the line segments are assigned default serial numbers in the order in which they are formed, and connected in the order of the default serial numbers, and finally the line segments can be connected as follows: Figure 6 The indoor space area shown is the vectorized data structure of the indoor space.

[0144] Correspondingly, the embodiments of the present disclosure also provide device embodiments corresponding to the aforementioned method embodiments. Figure 7 FIG. 1 is a structural block diagram of a point cloud data vectorization device according to an embodiment of the present invention. Figure 7 As shown, a vectorization device for point cloud data may include a plane structure classification module 701, a first determination module 702, a second determination module 703, a structure completion module 704 and a graph cut processing module 705, wherein:

[0145] A plane structure classification module 701 is used to classify the plane structure of the target indoor space based on the point cloud data image of the target indoor space to obtain multiple categories of plane structures;

[0146] A first determining module 702 is used to determine a target plane structure from the multiple types of plane structures;

[0147] The second determination module 703 is used to determine the top view image of the point cloud data image of the target indoor space before performing the graph cut algorithm processing on the target plane structure to obtain the vectorized data structure corresponding to the point cloud data image of the target indoor space;

[0148] A structure completion module 704 is used to perform structure completion processing on all plane structures of the target indoor space in the top view image to obtain the completed plane structure of the target indoor space;

[0149] The graph cut processing module 705 is used to perform graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space, so as to express the topological structure of the target indoor space through the vectorized data structure.

[0150] In summary, based on the point cloud data image of the target indoor space, the plane structure of the target indoor space is classified, and then the target plane structure is determined from multiple categories of plane structures, and then the target plane structure is processed by the graph cutting algorithm to obtain the vectorized data structure corresponding to the point cloud data image of the target indoor space, thereby realizing a fully automated vectorization solution for the point cloud data image of the target indoor space, improving the efficiency of mapping and tracing the target indoor space, and ensuring the consistency of the measured dimensions, avoiding the measurement accuracy variance caused by manual measurement by the operator, and improving the accuracy of mapping and tracing the target indoor space. In addition, by performing structural completion processing on all plane structures of the target indoor space, a more complete plane structure can be obtained, further improving the accuracy of vectorization of point cloud data.

[0151] As an optional embodiment of the present disclosure, the structure completion module 704 may include: a plane adding unit, used to add a first supplementary plane perpendicular to the direction of the normal vector and the gravity direction of any planar structure in the overhead image, passing through any end point of the overhead line segment of any planar structure, wherein the first supplementary plane is used to perform structural completion on any planar structure.

[0152] As an optional embodiment of the present disclosure, the structure completion module 704 may further include: a stretching unit, used to stretch the contour line segments of any planar structure in the overhead image according to a preset height value in the direction of the target coordinate axis, so as to obtain a second supplementary plane corresponding to the contour line segments of any planar structure, wherein the second supplementary plane is used to structurally supplement the planar structure corresponding to the contour line segments of any planar structure.

[0153] As an optional embodiment of the present disclosure, the first determining module 702 may include: a labeling unit, configured to label all plane structures of the multiple categories, wherein the plane structures labeled with the first category label are target plane structures.

[0154] As an optional embodiment of the present disclosure, the graph cut processing module 705 may include: a solution space determination unit, used to determine the solution space constituted by the target plane structure based on the plane structure of the target indoor space after completion, the solution space being the image to be segmented processed by the graph cut algorithm, and used to solve the cut points and cut edges processed by the graph cut algorithm; a vectorized data determination unit, used to determine the vectorized data structure corresponding to the point cloud data image of the target indoor space in the solution space by using the visibility information of the point cloud data of the target indoor space.

[0155] As an optional embodiment of the present disclosure, the solution space determination unit may include: a first determination subunit, used to determine a first outer rectangular contour based on an outer bounding box of the point cloud data of the target indoor space; a second determination subunit, used to determine a target plane structure from the completed plane structure of the target indoor space; and a third determination subunit, used to determine a solution space constituted by the target plane structure based on the first outer rectangular contour and the target plane structure.

[0156] As an optional embodiment of the present disclosure, the third determination subunit can be further used to: expand and adjust the first outer rectangular contour according to a preset size to obtain a second outer rectangular contour; determine the solution space constituted by the target plane structure based on the second outer rectangular contour and the target plane structure; wherein the first outer rectangular contour and the second outer rectangular contour are both rectangular polygons.

[0157] As an optional embodiment of the present disclosure, the third determination subunit, which implements the determination of the solution space constituted by the target plane structure based on the second outer rectangular contour and the target plane structure, may specifically include: determining the top view projection of the target plane structure to obtain a two-dimensional line segment corresponding to each of the target plane structures; inserting the two-dimensional line segment corresponding to each of the target plane structures into the rectangular polygon constituted by the second outer rectangular contour to divide the rectangular polygon constituted by the second outer rectangular contour into a plurality of rectangular polygonal areas, and the plurality of rectangular polygonal areas constitute the solution space.

[0158] As an optional embodiment of the present disclosure, the third determining subunit, which implements the inserting of the two-dimensional line segments corresponding to each of the target plane structures into the rectangular polygon formed by the second outer rectangular contour, so as to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygonal areas, may specifically include: determining the length value of the two-dimensional line segments corresponding to each of the target plane structures; and inserting the two-dimensional line segments into the rectangular polygon formed by the second outer rectangular contour in order from long to short based on the length values ​​of the two-dimensional line segments corresponding to all of the target plane structures, so as to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygonal areas.

[0159] As an optional embodiment of the present disclosure, the vectorized data determination unit may include: a fourth determination subunit, used to determine the coordinate position of the image acquisition device in the solution space, the image acquisition device being used to acquire point cloud data of the target indoor space; a connection subunit, used to connect the coordinate position of the image acquisition device to any point of the point cloud data of the target indoor space to obtain multiple connecting line segments; a weight setting subunit, used to set weights for multiple rectangular polygonal areas in the solution space passed by each of the connecting line segments according to a preset weight setting method; an area division subunit, used to divide the multiple rectangular polygonal areas in the solution space into indoor space and outdoor space according to the weight settings of the rectangular polygonal areas in the solution space; a vectorized data determination subunit, used to sequentially connect the rectangular polygonal areas of the indoor space and the outdoor space to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space.

[0160] As an optional embodiment of the present disclosure, the weight setting subunit can be further used to: determine the vertex of the graph corresponding to the rectangular polygonal area where the image acquisition device is located as the first vertex; determine the edge obtained by connecting the first vertex to the starting point determined by the graph cut algorithm as the first target edge; add a fixed weight value on the basis of the original weight value of the first target edge to obtain the final weight value of the first target edge; for any one of the multiple connecting line segments, determine that the multiple common edges involved in the rectangular polygonal area passed by the connecting line segment are the second target edges; add the fixed weight value on the basis of the original weight value of the second target edge to obtain the final weight value of the second target edge; determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which the point cloud data of the target indoor space in the solution space belongs; determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which any point cloud data belongs in the solution space as the third target edge; subtract the fixed weight value from the original weight value of the third target edge to obtain the final weight value of the third target edge.

[0161] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0162] Below, reference Figure 8 The electronic device according to the embodiment of the present disclosure is described. The electronic device may be any one or both of the first device and the second device, or a stand-alone device independent of them, and the stand-alone device may communicate with the first device and the second device to receive the collected input signals from them.

[0163] Figure 8 A block diagram of an electronic device according to an embodiment of the present disclosure is illustrated.

[0164] like Figure 8 As shown, the electronic device includes one or more processors and memory.

[0165] The processor may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0166] The memory may store one or more computer program products, and the memory may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program products may be stored on the computer-readable storage medium, and the processor may run the computer program product to implement the vectorization method of point cloud data of the various embodiments of the present disclosure described above and / or other desired functions.

[0167] In one example, the electronic device may further include: an input device and an output device, and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0168] In addition, the input device may also include, for example, a keyboard, a mouse, and the like.

[0169] The output device can output various information to the outside, including determined distance information, direction information, etc. The output device can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0170] Of course, to simplify, Figure 8 Only some of the components related to the present disclosure in the electronic device are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device may further include any other appropriate components.

[0171] In addition to the above-mentioned methods and devices, an embodiment of the present disclosure may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps in the vectorization method of point cloud data according to various embodiments of the present disclosure described in the above part of this specification.

[0172] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0173] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the steps of the method for vectorizing point cloud data according to various embodiments of the present disclosure described in the above part of this specification.

[0174] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0175] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, and are not limitations. The above details do not limit the present disclosure to the necessity of adopting the above specific details to be implemented.

[0176] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0177] The block diagrams of the devices, apparatuses, equipment, and systems involved in this disclosure are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including," "comprising," "having," and the like are open words, referring to "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or," and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0178] The method and apparatus of the present disclosure may be implemented in many ways. For example, the method and apparatus of the present disclosure may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present disclosure are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present disclosure may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present disclosure. Therefore, the present disclosure also covers a recording medium storing a program for executing the method according to the present disclosure.

[0179] It should also be noted that in the apparatus, device and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0180] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0181] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A method for vectorizing point cloud data, characterized in that: The method comprises: Based on the point cloud data image of the target indoor space, the plane structure of the target indoor space is classified to obtain plane structures of multiple categories; Determine a target plane structure from the plane structures of the plurality of categories; Determine a top view image of the point cloud data image of the target indoor space; In the overhead image, performing structural completion processing on all plane structures of the target indoor space to obtain the completed plane structure of the target indoor space; The target plane structure is processed by a graph cut algorithm to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space, so as to express the topological structure of the target indoor space through the vectorized data structure.

2. The method according to claim 1, characterized in that The step of determining a target plane structure from the plane structures of the plurality of categories comprises: All plane structures of the multiple categories are labeled, wherein the plane structures labeled with the first category label are target plane structures.

3. The method according to claim 1 or 2, characterized in that: The step of performing graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space includes: Based on the plane structure of the target indoor space after the completion, determining a solution space formed by the target plane structure, wherein the solution space is the image to be segmented processed by the graph cut algorithm, and is used to solve the cut points and cut edges processed by the graph cut algorithm; In the solution space, visibility information of the point cloud data of the target indoor space is used to determine a vectorized data structure corresponding to the point cloud data image of the target indoor space.

4. The method according to claim 3, characterized in that The determining, based on the completed plane structure of the target indoor space, a solution space formed by the target plane structure comprises: Determine a first outer rectangular contour according to an outer bounding box of the point cloud data of the target indoor space; Determining a target plane structure from the completed plane structure of the target indoor space; A solution space formed by the target plane structure is determined based on the first outer rectangular outline and the target plane structure.

5. The method according to claim 4, characterized in that The determining of a solution space formed by the target plane structure based on the first outer rectangular contour and the target plane structure includes: Enlarging and adjusting the first outer rectangular outline according to a preset size to obtain a second outer rectangular outline; Determining a solution space formed by the target plane structure based on the second outer rectangular contour and the target plane structure; Wherein, the first outer rectangular outline and the second outer rectangular outline are both rectangular polygons.

6. The method according to claim 5, characterized in that The determining of a solution space formed by the target plane structure based on the second outer rectangular contour and the target plane structure includes: Determine a top-down projection diagram of the target plane structure, and obtain a two-dimensional line segment corresponding to each of the target plane structures; Inserting the two-dimensional line segment corresponding to each of the target plane structures into the rectangular polygon formed by the second outer rectangular contour to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygon areas, wherein the plurality of rectangular polygon areas constitute the solution space.

7. The method according to claim 6, characterized in that The step of inserting the two-dimensional line segment corresponding to each of the target plane structures into the rectangular polygon formed by the second outer rectangular contour to divide the rectangular polygon formed by the second outer rectangular contour into a plurality of rectangular polygon areas includes: Determine the length value of the two-dimensional line segment corresponding to each of the target plane structures; Based on the length values ​​of the two-dimensional line segments corresponding to all the target plane structures, the two-dimensional line segments are inserted into the rectangular polygon formed by the second outer rectangular outline in order from long to short, so as to divide the rectangular polygon formed by the second outer rectangular outline into multiple rectangular polygon areas.

8. The method according to any one of claims 4 to 7, characterized in that: The method of determining, in the solution space, a vectorized data structure corresponding to the point cloud data image of the target indoor space by using visibility information of the point cloud data of the target indoor space comprises: Determining the coordinate position of an image acquisition device in the solution space, wherein the image acquisition device is used to acquire point cloud data of the target indoor space; Connecting the coordinate position of the image acquisition device to any point of the point cloud data of the target indoor space to obtain a plurality of connecting line segments; Setting weights for a plurality of rectangular polygonal areas in the solution space through which each of the connecting line segments passes according to a preset weight setting method; Dividing a plurality of rectangular polygonal areas in the solution space into indoor space and outdoor space according to weight settings of the rectangular polygonal areas in the solution space; The rectangular polygonal areas of the indoor space and the outdoor space are sequentially connected to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space.

9. The method according to claim 8, characterized in that The step of setting weights for the plurality of rectangular polygonal areas in the solution space through which each connecting line segment passes according to a preset weight setting method includes: Determine the vertex of the graph corresponding to the rectangular polygonal area where the image acquisition device is located as the first vertex; Determine an edge obtained by connecting the first vertex to the starting point determined by the graph cut algorithm as a first target edge; Adding a fixed weight value to an original weight value of the first target edge to obtain a final weight value of the first target edge; For any one of the plurality of connecting line segments, determining a plurality of common edges involved in a rectangular polygonal area passed through by the connecting line segment as second target edges; Adding the fixed weight value to the original weight value of the second target edge to obtain a final weight value of the second target edge; Determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which the point cloud data of the target indoor space in the solution space belongs; Determine the edge of the graph cut algorithm corresponding to the rectangular polygonal area to which any point cloud data belongs in the solution space as the third target edge; The fixed weight value is subtracted from the original weight value of the third target edge to obtain a final weight value of the third target edge.

10. A point cloud data vectorization device, characterized in that: The device comprises: A plane structure classification module is used to classify the plane structure of the target indoor space based on the point cloud data image of the target indoor space to obtain plane structures of multiple categories; A first determining module, configured to determine a target plane structure from the plane structures of the plurality of categories; A second determination module is used to determine a top view image of the point cloud data image of the target indoor space before performing the graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space; A structure completion module, used for performing structure completion processing on all plane structures of the target indoor space in the top view image to obtain the completed plane structure of the target indoor space; The graph cut processing module is used to perform graph cut algorithm processing on the target plane structure to obtain a vectorized data structure corresponding to the point cloud data image of the target indoor space, so as to express the topological structure of the target indoor space through the vectorized data structure.

Images