CN-122023508-A - Point cloud deviation rectifying method, system, storage medium and equipment based on building information model

CN122023508ACN 122023508 ACN122023508 ACN 122023508ACN-122023508-A

Abstract

The invention discloses a point cloud deviation correcting method based on a building information model, which comprises the steps of obtaining original point cloud data to be processed and a corresponding building information model, dividing the original point cloud data into a plurality of local point cloud sub-blocks based on a space topological structure of the model, extracting features, obtaining main plane features of the local point cloud sub-blocks, registering the main plane features of the local point cloud sub-blocks with corresponding space areas and structural members in the building information model, obtaining target poses of each local point cloud sub-block, sampling in the original point cloud data to generate a plurality of nodes, constructing a deformation map covering the original point cloud data, constructing an energy function containing a rigid constraint item, a smoothing regularization item and a position constraint item by taking the target poses of each local point cloud sub-block as position constraints, minimizing the energy function, solving optimal transformation parameters of the plurality of nodes in the deformation map, and calculating the final position of each point in the original point cloud data based on the parameters, so as to generate the corrected point cloud.

Inventors

ZHU YUANBIAO
ZU YANAN

Assignees

绘见科技(深圳)有限公司

Dates

Publication Date: 20260512
Application Date: 20260203

Claims (10)

1. The point cloud deviation rectifying method based on the building information model is characterized by comprising the following steps of: Acquiring original point cloud data to be processed and a corresponding building information model; dividing the original point cloud data into a plurality of local point cloud sub-blocks based on the spatial topological structure of the building information model; extracting the characteristics of each local point cloud sub-block to obtain the main plane characteristics of the local point cloud sub-block; Registering the main plane characteristics of the local point cloud sub-blocks with corresponding space areas and structural members in the building information model to obtain target pose of each local point cloud sub-block; Uniformly sampling in the original point cloud data to generate a plurality of nodes, and constructing a deformation graph covering the original point cloud data; Constructing an energy function comprising a rigid constraint term, a smooth regularization term and a position constraint term by taking the target pose of each local point cloud sub-block as a position constraint; minimizing the energy function, and solving optimal transformation parameters of a plurality of nodes in the deformation graph; And calculating the final position of each point in the original point cloud data based on the optimal transformation parameters of the plurality of nodes, and generating the corrected point cloud.
2. The method for correcting a point cloud based on a building information model according to claim 1, wherein the method for dividing the original point cloud data into a plurality of local point cloud sub-blocks based on a spatial topology structure of the building information model specifically comprises: Taking a wall body and a floor slab in the building information model as physical boundaries, and determining a segmentation area by taking a room or an axis network as a basic unit; Cutting the original point cloud data into a plurality of local sub-blocks; and setting a buffer zone with a preset width at the cutting boundary of the adjacent units, so that the adjacent partial point cloud sub-blocks contain partial same point cloud data, wherein the width of the buffer zone is determined according to the estimated accumulated drift amount and the partial characteristic geometric scale.
3. The method for correcting the point cloud based on the building information model according to claim 1, wherein registering the main plane characteristics of the local point cloud sub-blocks with the corresponding spatial regions and structural members in the building information model to obtain the target pose of each local point cloud sub-block specifically comprises: Acquiring a grid surface of a structural member in a space area corresponding to the local point cloud sub-block in the building information model; And constructing an error function based on the point-to-plane distance, and solving a translation vector of the local point cloud sub-block in the building information model and a rotation matrix representing three-dimensional space orientation by iteratively minimizing the distance between the main plane characteristic of the local point cloud sub-block and the grid surface of the structural member, wherein the translation vector and the rotation matrix represent the three-dimensional space orientation and serve as target pose.
4. The method for correcting the point cloud based on the building information model according to claim 3, wherein the constructing an energy function including a rigid constraint term, a smooth regularization term and a position constraint term by using the target pose of each local point cloud sub-block as a position constraint specifically includes: Wherein, the Is a rigid constraint term; Smoothing the term for regularization; Is a position constraint item; 、、 The weight coefficients of the corresponding items are respectively.
5. The building information model-based point cloud rectification method of claim 4, wherein a rigidity constraint term, a smoothing regularization term and a position constraint term are determined according to the following formula: the rigid constraint term is: Wherein, the For the rotation matrix of the j-th node in the deformation graph, The unit matrix is M, and M is the total number of nodes in the deformation graph; The smoothing regularization term is: Wherein, the And Initial coordinates of neighboring nodes j and k, t j and t k are translation vectors of neighboring nodes j and k, respectively, N (j) is a neighborhood node set of node j, Is a first weight coefficient; The position constraint item is: Wherein, the Is the three-dimensional coordinate of the ith point in the point cloud after deviation correction, The correct three-dimensional coordinates in the building information model corresponding to the point i in the point cloud are obtained; Obtaining the correct three-dimensional coordinates of the point cloud point i in the building information model according to the initial coordinates vi of the point cloud point i by applying target pose transformation of the local point cloud sub-block to which the point cloud point i belongs 。
6. The method for correcting a point cloud based on a building information model according to claim 5, wherein the minimizing the energy function and solving the optimal transformation parameters of a plurality of nodes in the deformation graph specifically comprise: initial transformation parameters are given to all nodes j in the deformation graph, namely a rotation matrix Translation vector Wherein the rotation matrix Translation vector The initial values of (1) are an identity matrix and a zero vector; Iterating the energy function, in the kth iteration, based on the current rotation matrix Translation vector Calculating an energy function and updating the transformation parameters based on the parameter update amounts Δr and Δt: Wherein, the 、 Respectively representing a rotation matrix and a translation vector of the kth iteration of the node j; 、 respectively representing a rotation matrix and a translation vector of the (k+1) th iteration of the node j, wherein DeltaR and Deltat respectively represent corresponding parameter updating amounts; When the energy difference value of two adjacent iterations of the energy function is smaller than a preset threshold value, ending the iteration, and obtaining an optimal rotation matrix And an optimal translation vector As the optimal transformation parameters for node j.
7. The method for rectifying a point cloud based on a building information model according to claim 6, wherein calculating a final position of each point in the original point cloud data based on the optimal transformation parameters of the plurality of nodes, and generating a rectified point cloud specifically comprises: For each point in the raw point cloud data Determining K nearest neighbor nodes affecting the point to form a node set ; Calculate the point To adjacent nodes Is of the interpolation weights of (1) The weight is inversely proportional to the distance from the point to the node, and satisfies: ; Node-based Is an optimal rotation matrix of (a) And an optimal translation vector The final position of the point is calculated by : Wherein, the Is taken as a point K neighbor node sets of (a); Is taken as a point To adjacent nodes Is used for the interpolation weight of the (a); And Is a node Is an optimal rotation matrix of (a) And an optimal translation vector ; The initial position coordinates of the j-th node.
8. The point cloud deviation rectifying system based on the building information model is characterized by comprising a data acquisition unit, a feature extraction unit, a registration unit, a deformation graph construction unit, an energy function construction unit, a solving unit and a deviation rectifying generation unit; The data acquisition unit is used for acquiring the original point cloud data to be processed and the corresponding building information model; The feature extraction unit is used for dividing the original point cloud data into a plurality of local point cloud sub-blocks based on the space topological structure of the building information model; The registration unit is used for registering the main plane characteristics of the local point cloud sub-blocks with the corresponding space areas and structural members in the building information model to obtain the target pose of each local point cloud sub-block; the deformation graph construction unit is used for uniformly sampling in the original point cloud data to generate a plurality of nodes and constructing a deformation graph covering the original point cloud data; The energy function construction unit is used for constructing an energy function comprising a rigid constraint term, a smooth regularization term and a position constraint term by taking the target pose of each local point cloud sub-block as a position constraint; the solving unit is used for minimizing the energy function and solving the optimal transformation parameters of a plurality of nodes in the deformation graph; And the deviation rectifying generation unit is used for calculating the final position of each point in the original point cloud data based on the optimal transformation parameters of the plurality of nodes and generating a point cloud after deviation rectifying.
9. A readable storage medium storing a computer program, which when executed by a processor causes the processor to perform the steps of the method according to any one of claims 1 to 7.
10. A computer device comprising a memory and a processor, wherein the memory stores a computer program which, when executed by the processor, causes the processor to perform the steps of the method as claimed in any one of claims 1 to 7.

Description

Point cloud deviation rectifying method, system, storage medium and equipment based on building information model Technical Field The invention relates to the technical field of data processing, in particular to a point cloud deviation rectifying method, a system, a storage medium and equipment based on a building information model. Background With the wide application of the three-dimensional laser scanning technology in the fields of construction, civil engineering and the like, the large scene point cloud data obtained through the synchronous positioning and map building (SLAM) technology becomes an important foundation for digital twinning, construction quality detection and the like. However, when the SLAM technique is operated in a scene with long distance, weak texture or closed loop missing, a cumulative error (Drift) is inevitably generated, so that the finally generated three-dimensional point cloud model has geometrical distortion phenomena such as integral bending, twisting or dislocation. The drift problem seriously affects the absolute precision of the point cloud data and the alignment effect of the point cloud data and a design model, and limits the direct application of the point cloud data in high-standard requirement scenes such as accurate measurement, reverse modeling and the like. In the prior art, the correction method for the point cloud drift is mainly divided into two types, namely a global optimization method based on closed loop detection, wherein the method depends on whether an effective closed loop exists in a scene, has limited effect in the scene which lacks obvious closed loop characteristics such as a long corridor and an underground space, and the like, and the correction method based on an external positioning reference (such as a GPS) cannot be used in an environment with serious indoor or satellite signal shielding. In recent years, researchers try to perform point cloud registration by using a Building Information Model (BIM) as a reference standard, but the existing method mostly adopts a global rigid registration strategy, and cannot effectively process nonlinear and non-rigid characteristics of SLAM drift. When there is a serious distortion in the point cloud, the global rigid registration can destroy the local geometry, generate huge registration errors in the drifting serious region, and even cause registration failure. Disclosure of Invention Based on the above, it is necessary to provide a point cloud correction method based on a building information model. A point cloud deviation rectifying method based on a building information model, the method comprising the following steps: Acquiring original point cloud data to be processed and a corresponding building information model; dividing the original point cloud data into a plurality of local point cloud sub-blocks based on the spatial topological structure of the building information model; extracting the characteristics of each local point cloud sub-block to obtain the main plane characteristics of the local point cloud sub-block; Registering the main plane characteristics of the local point cloud sub-blocks with corresponding space areas and structural members in the building information model to obtain target pose of each local point cloud sub-block; Uniformly sampling in the original point cloud data to generate a plurality of nodes, and constructing a deformation graph covering the original point cloud data; Constructing an energy function comprising a rigid constraint term, a smooth regularization term and a position constraint term by taking the target pose of each local point cloud sub-block as a position constraint; minimizing the energy function, and solving optimal transformation parameters of a plurality of nodes in the deformation graph; And calculating the final position of each point in the original point cloud data based on the optimal transformation parameters of the plurality of nodes, and generating the corrected point cloud. In the above solution, the dividing the original point cloud data into a plurality of local point cloud sub-blocks based on the spatial topology structure of the building information model specifically includes: Taking a wall body and a floor slab in the building information model as physical boundaries, and determining a segmentation area by taking a room or an axis network as a basic unit; Cutting the original point cloud data into a plurality of local sub-blocks; and setting a buffer zone with a preset width at the cutting boundary of the adjacent units, so that the adjacent partial point cloud sub-blocks contain partial same point cloud data, wherein the width of the buffer zone is determined according to the estimated accumulated drift amount and the partial characteristic geometric scale. In the above scheme, registering the main plane feature of the local point cloud sub-block with a corresponding spatial region and a structural member in the building information m