CN113593017A - Method, device and equipment for constructing surface three-dimensional model of strip mine and storage medium - Google Patents
Method, device and equipment for constructing surface three-dimensional model of strip mine and storage medium Download PDFInfo
- Publication number
- CN113593017A CN113593017A CN202110879529.5A CN202110879529A CN113593017A CN 113593017 A CN113593017 A CN 113593017A CN 202110879529 A CN202110879529 A CN 202110879529A CN 113593017 A CN113593017 A CN 113593017A
- Authority
- CN
- China
- Prior art keywords
- data
- point cloud
- dimensional model
- fusion
- fusion processing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/251—Fusion techniques of input or preprocessed data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4007—Interpolation-based scaling, e.g. bilinear interpolation
Abstract
The invention discloses a method, a device, equipment and a storage medium for constructing a three-dimensional model of an earth surface of an open-pit mine. The method comprises the following steps: performing first fusion processing on local point cloud data acquired by acquisition equipment aiming at a target area based on a time dimension; performing second fusion processing on the local point cloud data based on the space dimension; performing target segmentation processing based on the data after the first fusion processing and the data after the second fusion processing; and updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine. Non-physical objects can be separated and the shielded local stope model can be restored, so that local point cloud fusion under multiple time and space conditions is realized, a real geographic space is restored, and a strip mine mining scene three-dimensional model is quickly constructed; in addition, the method can quickly and accurately construct the three-dimensional model of the open-pit mining scene, and can meet the requirements of the open-pit intelligent mining scene with frequently updated scenes.
Description
Technical Field
The invention relates to the field of strip mine exploitation, in particular to a method, a device, equipment and a storage medium for constructing a three-dimensional model of an earth surface of a strip mine.
Background
The three-dimensional automatic modeling of the open stope is an important prerequisite for open-pit intelligent mining. The existing modeling methods include: firstly, modeling is carried out based on a traditional measuring means, measuring equipment such as a total station is used for measuring a strip mine mining scene, measuring data are processed to obtain spatial discrete points on a slope top bottom, a tunnel boundary and the like, then an in-pit three-dimensional digital model (TIN) is generated based on constraint Delaunay, and then spatial Boolean operation is carried out on the TIN and a Digital Elevation Model (DEM) around a stope to generate a final strip mine three-dimensional digital model, and the method is difficult in data acquisition and low in modeling timeliness; secondly, modeling is carried out based on unmanned aerial vehicle oblique photogrammetry, a low-cost portable aerial camera is utilized to conveniently and quickly obtain a strip mine photo sequence, a motion and structure reconstruction method based on an image sequence is used for automatically extracting three-dimensional feature points of a stope, generated point cloud data of each area of the strip mine are automatically fused into a complete three-dimensional model based on a beam adjustment method, the later modeling calculation amount is huge, the modeling time is long, meanwhile, the elimination of non-geographic objects such as equipment in a scene is difficult, and the robustness of the model needs to be further improved; and thirdly, modeling is carried out based on a laser scanner, data acquisition is carried out in a static state, three-dimensional laser point cloud data are acquired through the scanner, high-precision earth surface three-dimensional measurement data under the space coordinate reference are acquired through combined resolving of various data by combining absolute coordinate information of RTK-GNSS, but the data acquisition period is longer, the processing of some special terrains seriously sheltered is poorer, the later-stage data processing is time-consuming, and the timeliness is low.
In summary, the existing modeling method cannot be applied to an open-air intelligent mining scene with frequent environment updating, cannot meet a high-precision three-dimensional terrain model required by intelligent mine operation, and has the problems of high data acquisition cost, low resolution, no separation of geographic and non-geographic objects in the model and the like.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method, an apparatus, a device and a storage medium for constructing a three-dimensional model of a surface of a strip mine, which aim to separate a non-physical object and restore a blocked local stope model, quickly construct a three-dimensional model of a mining scene of the strip mine, and meet mining requirements of the strip mine.
The technical scheme of the embodiment of the invention is realized as follows:
in a first aspect, an embodiment of the present invention provides a method for building a three-dimensional model of a surface of an open pit mine, including:
performing first fusion processing on local point cloud data acquired by acquisition equipment aiming at a target area based on a time dimension;
performing second fusion processing on the local point cloud data based on the spatial dimension;
performing target segmentation processing based on the data after the first fusion processing and the data after the second fusion processing;
and updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine.
In the foregoing aspect, before the first fusion process and the second fusion process, the method further includes:
acquiring local point cloud data acquired by acquisition equipment aiming at a target area;
performing data preprocessing on the local point cloud data, wherein the data preprocessing comprises the following steps: point cloud grouping, time sequence serialization and data deletion processing;
correspondingly, the first fusion processing and the second fusion processing are carried out on the basis of the local point cloud data after the data preprocessing.
In the foregoing solution, the performing a first fusion process on the local point cloud data acquired by the acquisition device for the target area based on the time dimension includes:
acquiring an attitude transformation matrix between adjacent frame point clouds;
and performing interframe matching on the local point cloud data based on the attitude transformation matrix to obtain the data after the first fusion processing.
In the above scheme, the method further comprises:
determining the attitude transformation matrix for point cloud data at different moments in the local point cloud data based on linear interpolation transformation; alternatively, the first and second electrodes may be,
the attitude transformation matrix is determined based on detection data of an inertial measurement unit.
In the foregoing solution, the performing a second fusion process on the local point cloud data based on the spatial dimension includes:
carrying out rough matching on the point cloud data in the local point cloud data based on the positioning information of the acquisition equipment;
and carrying out spatial fusion on the roughly matched local point cloud data based on feature identification to obtain data after second fusion processing.
In the foregoing solution, the performing target segmentation processing based on the data after the first fusion processing and the data after the second fusion processing includes:
carrying out segmentation processing on the geographic object and the non-geographic object on the data subjected to the first fusion processing and the data subjected to the second fusion processing to obtain data from which the non-geographic object is deleted;
wherein the geographic object comprises at least one of: a slope surface and a step surface; the non-geographic object includes at least one of: transport trucks, excavators and bulldozers.
In the foregoing solution, the updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine includes:
and fusing the data after the non-geographic object is deleted with global point cloud data corresponding to the historical three-dimensional model of the surface of the strip mine, and determining an updated three-dimensional model based on the fused new global point cloud data.
In a second aspect, an embodiment of the present invention further provides an apparatus for building a three-dimensional model of a surface of a strip mine, including:
the first fusion processing module is used for performing first fusion processing on local point cloud data acquired by the acquisition equipment aiming at the target area based on the time dimension;
the second fusion processing module is used for carrying out second fusion processing on the local point cloud data based on the space dimension;
the segmentation processing module is used for carrying out target segmentation processing on the basis of the data subjected to the first fusion processing and the data subjected to the second fusion processing;
and the model updating module is used for updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine.
In a third aspect, an embodiment of the present invention further provides an apparatus for building a three-dimensional model of a surface of an open pit mine, including: a processor and a memory for storing a computer program capable of running on the processor, wherein the processor, when running the computer program, is configured to perform the steps of the method according to an embodiment of the invention.
In a fourth aspect, an embodiment of the present invention further provides a storage medium, where a computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the method according to the embodiment of the present invention are implemented.
According to the technical scheme provided by the embodiment of the invention, local point cloud data acquired by acquisition equipment aiming at a target area is subjected to first fusion processing based on a time dimension; performing second fusion processing on the local point cloud data based on the space dimension; performing target segmentation processing based on the data after the first fusion processing and the data after the second fusion processing; and updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine. Non-physical objects can be separated and the shielded local stope model can be restored, so that fusion of local point clouds under multiple time and space is realized, a real geographic space is restored, and a three-dimensional model of a mining scene of the strip mine is quickly constructed; in addition, the method can quickly and accurately construct the three-dimensional model of the surface mining scene, solves the problems of long modeling period, complex steps, low modeling timeliness and the like of the traditional modeling technology, and can meet the requirements of the surface intelligent mining scene with frequently updated scenes.
Drawings
FIG. 1 is a schematic flow chart of a method for constructing a three-dimensional model of the surface of a strip mine according to an embodiment of the invention;
FIG. 2 is a schematic structural diagram of an acquisition device in an exemplary application of the present invention;
FIG. 3 is a top view of FIG. 2;
FIG. 4 is a schematic view of a surface mine scenario in an example of an application of the present invention;
FIG. 5 is a schematic diagram of a global point cloud generated after matching fusion according to an exemplary embodiment of the present invention;
FIG. 6 is a partial cloud plot of laser radar generated in an exemplary application of the present invention;
FIG. 7 is a schematic structural diagram of a surface three-dimensional model construction device for a strip mine according to an embodiment of the invention;
FIG. 8 is a schematic structural diagram of a three-dimensional surface model building device for a strip mine according to an embodiment of the invention;
fig. 9 is a schematic structural diagram of a three-dimensional model building system for a surface of a strip mine according to an embodiment of the invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
Before further detailed description of the embodiments of the present invention, terms and expressions mentioned in the embodiments of the present invention are explained, and the terms and expressions mentioned in the embodiments of the present invention are applicable to the following explanations:
in the related technology, the traditional data measuring equipment has long working period and complex modeling process, so that the requirement of dynamic change of the working scene of the strip mine cannot be met; the unmanned aerial vehicle oblique photography modeling is huge in calculated amount, non-geographic objects in a scene can not be removed almost, and the robustness needs to be further improved; the single-acquisition-point vehicle-mounted laser radar modeling is long in data acquisition period, difficult in recovery of seriously sheltered terrain, and difficult to perform the task of efficient and accurate modeling of an open-pit mining scene.
Based on the above, in various embodiments of the invention, deep fusion of the local point clouds in two dimensions of space and time is carried out, meanwhile, target segmentation is carried out on equipment and structure objects by combining the distribution characteristics of the equipment and the structures in a mining scene, point cloud distortion caused by a mobile carrier and a mobile target is solved through motion compensation, confidence calculation is further carried out on the segmented and fused local point clouds, finally, the local point clouds formed in the modeling period are matched and fused with the global point clouds established in history to form the global point clouds in the period, three-dimensional reconstruction is carried out on the global point clouds to form a non-geographic object separation but geometrically seamless open-air mining scene three-dimensional vector model, and the timeliness and the accuracy of modeling are effectively improved.
As shown in fig. 1, an embodiment of the present invention provides a method for building a three-dimensional model of a surface of an open pit mine, including:
102, performing second fusion processing on the local point cloud data based on the space dimension;
103, performing target segmentation processing based on the data after the first fusion processing and the data after the second fusion processing;
and 104, updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine.
It can be understood that, in the embodiment of the present invention, based on the first fusion processing corresponding to the time dimension and the second fusion processing corresponding to the space dimension, and performing the target segmentation processing on the data after the first fusion processing and the data after the second fusion processing, the non-physical object can be separated and the blocked local stope model can be restored, so that the fusion of the local point clouds in multiple time and space is realized, the real geographic space is restored, and the three-dimensional model of the mining scene of the strip mine is quickly constructed; in addition, the method can quickly and accurately construct the three-dimensional model of the surface mining scene, solves the problems of long modeling period, complex steps, low modeling timeliness and the like of the traditional modeling technology, and can meet the requirements of the surface intelligent mining scene with frequently updated scenes.
It is to be understood that, for the three-dimensional modeling of the surface of the strip mine, since the strip mine body and the stope and other surrounding structures are interrelated, the target area may be at least one relevant area of the surface of the strip mine, which is not limited by the embodiment of the present application.
Illustratively, before the first fusing process and the second fusing process, the method further comprises:
acquiring local point cloud data acquired by acquisition equipment aiming at a target area;
performing data preprocessing on the local point cloud data, wherein the data preprocessing comprises the following steps: point cloud grouping, time sequence serialization and data deletion processing;
correspondingly, the first fusion processing and the second fusion processing are carried out on the basis of the local point cloud data after the data preprocessing.
The collection device may illustratively be a mobile device of an open pit mine, such as an electric shovel or mine card, with various collection sensors installed. As shown in fig. 2 and 3, three laser radar (LIDAR)1, two high-precision real-time positioning systems (RTK-GNSS)2 and an Inertial Measurement Unit (IMU)3 may be disposed on the mine card. It should be noted that the types and numbers of the above-mentioned acquisition sensors are only examples. Those skilled in the art can perform reasonable setting and adjustment according to needs, and the embodiment of the present application is not limited thereto.
In practical application, sensors such as laser radars and the like on each acquisition device can be calibrated, so that each laser radar is time-synchronized, data fusion is facilitated, errors are reduced, and a data interface is developed to acquire point cloud data in real time; when each acquisition device operates, point cloud data acquired by different acquisition devices and different spatial positions are subjected to point cloud grouping and are serialized in time sequence.
For example, the collected point cloud data may exceed the data amount required for modeling, and according to different sources of the point cloud data, for example, the quality of some point clouds collected by a laser radar is higher, the error of the point clouds collected in heavy rain or heavy fog weather is often very large, and the quality of the point clouds collected by a scraper is not as good as that of a mine truck due to dust, the point clouds are subjected to prior constraint (for example, a standard for determining how many point clouds are finally selected under different conditions) or are compared and analyzed with the high-quality point clouds, and partial low-quality point clouds are sampled or discarded directly, so that data deletion processing on the point cloud data is realized.
Illustratively, the data pruning process may be based on error analysis, comparing the point cloud data under poor conditions with the point cloud data under relatively good conditions to obtain a deviation value (i.e., error) of the point cloud under poor conditions, discarding or reducing the low quality point cloud data by down-sampling based on the deviation value obtained from historical error analysis, or based on criteria empirically determined for processing a large number of point clouds.
Illustratively, the performing a first fusion process on the local point cloud data acquired by the acquisition device for the target area based on the time dimension includes:
acquiring an attitude transformation matrix between adjacent frame point clouds;
and performing interframe matching on the local point cloud data based on the attitude transformation matrix to obtain the data after the first fusion processing.
Illustratively, the method further comprises:
determining the attitude transformation matrix for point cloud data at different moments in the local point cloud data based on linear interpolation transformation; alternatively, the first and second electrodes may be,
the attitude transformation matrix is determined based on detection data of an inertial measurement unit.
It can be understood that, in the time dimension, a method of inter-frame matching may be used, where there is a time relationship between adjacent frames scanned by the laser radar each time, which is also a basis for constructing a local point cloud map, and in order to perform matching fusion on point clouds between adjacent frames, a posture transformation matrix between point clouds of adjacent frames is first found, so as to obtain a corresponding relationship between a certain point in a current frame and a point set in a previous frame, and a motion of the laser radar is estimated to eliminate motion distortion.
It should be noted that the attitude transformation matrix can be obtained from the point cloud relationship at different times through linear interpolation transformation, and can also be obtained through measurement by an IMU (inertial measurement unit), the former is suitable for a state where the radar is at a low speed and at a constant speed, and the latter is applicable to a wide range and has high accuracy, but also has a problem of cost increase, and in application, different methods can be adopted in consideration of actual conditions.
The local point cloud fusion method based on linear interpolation transformation will be described below.
The local point cloud fusion method based on linear interpolation transformation can be divided into two steps of feature point extraction and feature matching. The purpose of the first step of feature point extraction is to reduce the time consumption of calculation, avoid using all point clouds for processing, and use feature points to replace complete data frames, expect to use the curvature calculation method to extract feature points, and introduce a new feature evaluation criterion, i.e. local plane smoothness c, to describe the curvature condition of the scanning points, as formula (1):
in the formula (1), the first and second groups,
to define points in the radar coordinate system L, the set S is the set of points included in one laser scan. Calculating the curvature c of all the points in the laser scanning set S through the formula (1), comparing with a set threshold value, classifying the points which are larger than the set threshold value as edge points and smaller than the set threshold valueOr the points equal to the set threshold are classified into plane points, so that an original point cloud set is divided into an edge point set and a plane point set, and the feature points in one frame are extracted; the second step of feature matching aims to match the extracted feature points of the single frame with the frame, and then perform inter-frame motion estimation to complete the calculation of the radar odometer, which is defined as follows:
tkthe time stamp of the beginning of the k-th scanning is the original point cloud obtained by scanningkThe corresponding time stamp at the end of the scan is tk+1And is combined with PkReprojection of tk+1At the moment
The edge point set extracted from the image is recorded as
Set of plane points is noted
tk+1And Pk+1And so on.
Inter-frame matching, i.e. using
Edge line in (1) and Pk+1Edge points in (1) are matched and used
Point of plane in (1) and Pk+1The matching problem of the edge point is firstly explained, because the edge point is a point formed by lines in a three-dimensional structure, the problem is that the shortest distance from the point to the line is obtained, and the nearest search method of KD-Tree is used for searching from the point to the line
Taking out a point, recording as a point i, and recording as a coordinate
In that
Taking out a corner characteristic point closest to the point i, and recording the corner characteristic point as a point j and the coordinate as a point
In addition, to prevent the collinear condition, we also need to extract another corner feature point closest to the point i from the scanning line different from the point j, and record the feature point as a point l and the coordinate as a point
After finding out the corresponding relationship between the feature points j and l, the shortest distance d from the point i to the line lj can be calculated by adopting the formula (2)EFor subsequent motion estimation.
Likewise, for matching points to planes, from
Taking out a point, recording as a point i, and recording as a coordinate
Since the plane is represented by three points, it is necessary to do so from
Three non-collinear feature points are taken out for feature plane representation. In that
Taking a plane feature point nearest to the point i, recording the plane feature point as a point j, and recording the coordinate as a point j
In the same scan line at point jTaking another plane feature point nearest to the point i, and recording as a point l and a coordinate as a point
Finally, in another scanning line, a plane characteristic point which is most adjacent to the point i is taken out and recorded as a point m, and the coordinate is recorded as a point m
The shortest distance d between the point i and the plane ljm is calculated using equation (3)HFor subsequent motion estimation.
And performing motion estimation, wherein the method requires that the radar is in a low-speed and uniform motion state, so that an attitude transformation matrix of receiving points at different moments can be obtained through linear interpolation transformation, thereby obtaining the corresponding relation between a certain point in the current frame and a point set in the previous frame, estimating the motion of the radar, further removing motion distortion, and obtaining P in the previous stepk+1Some edge point in and plane points i and PkThe geometric relationship between the median edge line and the plane, from which the points i and P can be derivedkCoordinate transformation matrix of center point set
Then, the interval [ t ] is obtained according to the interpolation formula (4)k+1,t]Attitude transformation information of radar-added point cloud
Therefore, point cloud matching and fusion between frames are completed.
Illustratively, the second fusion processing of the local point cloud data based on the spatial dimension includes:
carrying out rough matching on the point cloud data in the local point cloud data based on the positioning information of the acquisition equipment;
and carrying out spatial fusion on the roughly matched local point cloud data based on feature identification to obtain data after second fusion processing.
It can be understood that the point cloud fusion and matching are mainly performed in two steps in the space dimension, and the point cloud fusion in the space dimension is realized by firstly performing rough matching according to the positioning condition when each device collects point cloud information and then matching the markers in the generated point cloud map according to feature identification.
Considering that the spatial point cloud fusion method and technology are still immature, the data after the second fusion processing can be used as the supplement of the data after the first fusion processing, the main purpose is to restore the sheltered open-pit mine scene as far as possible and improve the modeling efficiency, and the schematic diagram of restoring the open-pit scene through multi-source point cloud spatial fusion is shown in fig. 4. The fusion mode can be divided into two steps, firstly, because point Cloud data acquired under multiple viewing angles are too large and disordered, the point Cloud data are considered to be firstly primarily processed in a point Cloud library PCL (Point Cloud library), then, redundant and disordered point Cloud information is removed from massive point Cloud data by adopting a probability theory analysis or filtering method, the integrity of an open-air mining scene is ensured, secondly, a point Cloud model database is established for some typical structures, equipment and structures in an open-air mine, and matching is carried out under the existing model constraint condition, so that the success rate can be greatly improved.
Illustratively, the performing the target segmentation processing based on the first fusion processed data and the second fusion processed data includes:
carrying out segmentation processing on the geographic object and the non-geographic object on the data subjected to the first fusion processing and the data subjected to the second fusion processing to obtain data from which the non-geographic object is deleted;
wherein the geographic object comprises at least one of: a slope surface and a step surface; the non-geographic object includes at least one of: transport trucks, excavators and bulldozers.
It can be understood that after the point cloud data is subjected to fusion processing in the two dimensions of the time dimension and the space dimension, the target segmentation processing is performed in the next step. First, a rough segmentation of the geographic object and the non-geographic object is performed by projecting a frame of point cloud (i.e. the point cloud data after the first fusion process and the second fusion process) into an image, for example, setting a point cloud P obtained at time tt，piPoint cloud P at time ttProjecting the point cloud to a range image, obtaining the resolution of the projected image according to the scanning mode of the laser radar, the density of the point cloud and the scanning line in the vertical direction, then changing each three-dimensional space point into a two-dimensional space pixel point, and obtaining piEuclidean distance r from corresponding pixel point to sensoriAnd then listing and evaluating the distance image, roughly extracting the geographic object, and marking partial geographic objects in the scene, such as the ground, steps, working platforms and the like, by judging the characteristics of the vertical dimension because the vertical direction of the distance image also represents the characteristics of the vertical dimension in the original three-dimensional space. The tagged geographic object is not subsequently segmented. Then, the distance image is divided into a plurality of clusters by using an image-based division method, the point of the same cluster is marked with a unique identifier, the step can remove some tiny non-geographic objects (leaves, grass and the like) as noise points, reduce the interference caused by the non-repeated appearance of the tiny objects between adjacent frames, comprehensively consider the requirements of speed and precision, artificially set the number of data points of the category processed as the noise points, and remove the data category lower than the value as the noise points, so that the saved geographic and non-geographic objects are relatively obvious. The entire range image is divided into several relatively large categories and the points in each image are given the following information: segmentation labels, indices of rows and columns in the range image, and range value r of the sensoriTransmitting the residual points after the segmentation processing into a feature extraction module for subsequent feature extraction processing, and judging whether the points belong to the geographic object or not by calculating the vertical angle and judging a threshold value, and further judging whether the points belong to the geographic object or notThe method comprises the steps of separating geographic objects and non-geographic objects, carrying out clustering processing on the separated non-geographic objects, reserving corresponding clusters after setting judgment conditions, carrying out deep learning semantic segmentation, adopting a mode of 'target detection-target identification-target segmentation' to segment the non-geographic objects, carrying out feature extraction on object point clouds in combination with different visual angles, carrying out feature fusion, carrying out object identification in combination with deep learning under the prior constraint determined by the types of the non-geographic objects, establishing a model base and a feature base in combination with a specific scene of an open pit mine according to existing equipment and structures of the mine, and realizing separation of the non-geographic objects in a mode of 'target segmentation'. In this process, a semantic segmentation index IOU (cross-over ratio) or a pixel accuracy (pixel-accuracy) may also be considered for the second confidence calculation and error analysis.
It can be understood that, in the foregoing process, local point cloud fusion, error analysis and confidence calculation have been performed, and a non-geographic object is separated, and the last step is to match and fuse the formed local point cloud with the established global point cloud to form a new global point cloud.
Illustratively, the updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine includes:
and fusing the data after the non-geographic object is deleted with global point cloud data corresponding to the historical three-dimensional model of the surface of the strip mine, and determining an updated three-dimensional model based on the fused new global point cloud data.
Two methods for acquiring the global point cloud map are specifically described below, including two methods based on the sensor vision field and based on map optimization.
Based on two modes of map-to-map (map-to-map) fusion and frame-to-map (scan-to-map) of a sensor vision field, firstly, an algorithm (radar mileage calculation method) similar to the local point cloud fusion is used for extracting and matching feature points, and the difference is that the algorithm (mapping algorithm) is used for matching 10 times of the feature points, so that the consistency of map construction is ensured, and in the process, the local undistorted point cloud obtained by the previous fusion is continuously matched to the existing global map, so that the whole map is established, and the following definitions are firstly made:
{ W } is a global coordinate system coinciding with the radar coordinate system, the initial pose, where the coordinates of point i are expressed in the form of
At the end of the (k + 1) th scan, the undistorted point cloud generated by the mileage calculation method is
And correspondingly in the interval tk+1,tk+2]Point cloud pose transformation information on
Definition of QkProjection of all accumulated point clouds before k frames in a world coordinate system;
interval [ t ]k,tk+1]And (4) point cloud posture transformation information.
With the output of the radar mileage calculation method, the mapping algorithm will be
From tk+1Is delayed to tk+2Obtaining a corresponding attitude matrix
By passing
Can be combined with
Projected under the global coordinate system { W }, is recorded as
Finally pass through pair
Is continuously optimized, namely
Is matched to QkThe above. It should be noted that, in the process of extracting the feature points, although the extracting method is the same as the mileage calculation method, 10 times of point cloud feature points are used, and in order to improve the operation efficiency, the point cloud Q in the cubic space with the side length of 10m is usedkThe entire map point cloud is replaced; in the process of feature point matching, S' is set as QkThe selected feature point set in the cubic range distinguishes edge feature points and plane feature points in S ', calculates covariance matrix M of S', respectively records eigenvalue and eigenvector as V and E, uses the same shortest distance calculation method as in the mileage calculation method, and optimizes and solves
Will be provided with
Is matched to QkFinally, the coordinate of each point of the point cloud map in the mapping process is positioned under the current pose coordinate system of the radar, and after the mapping is completed, the coordinate of each point is positioned under the initial pose coordinate system of the radar, namely under the global coordinate system.
Based on graph optimization, firstly obtaining observation data of each node
Secondly, because the position estimation error of the laser point cloud mapping is low, and the situation that almost no error exists in a short period is assumed, a group of similar feature sets is selected to construct a corresponding global feature point cloud Qt-1。
Wherein k represents Qt-1Is then constructed by the LM optimization method
And Qt-1And finally, obtaining a final global map by using a loop detection method and utilizing gtsam optimization, as shown in fig. 5, wherein the block part is the built local point cloud map, as shown in fig. 6. And then, a complete strip mine three-dimensional geographic model can be constructed by combining a point cloud three-dimensional reconstruction technology. It should be noted that the three-dimensional model is only geometrically seamless, phenomena such as unreality may exist at the edge, and the three-dimensional model is further subjected to semantic segmentation when mine design data (such as step design parameters and the like) are used for point cloud fusion, so that the modeling precision can be improved, and the model is more fit to the reality.
In order to implement the method according to the embodiment of the present invention, an embodiment of the present invention further provides a three-dimensional model building apparatus for a surface of a strip mine, which is provided in a three-dimensional model building device for a surface of a strip mine, as shown in fig. 7, and the apparatus for building a three-dimensional model for a surface of a strip mine includes: a first fusion processing module 701, a second fusion processing module 702, a segmentation processing module 703 and a model updating module 704.
The first fusion processing module 701 is configured to perform first fusion processing on local point cloud data acquired by an acquisition device for a target region based on a time dimension; the second fusion processing module 702 is configured to perform a second fusion process on the local point cloud data based on the spatial dimension; the segmentation processing module 703 is configured to perform target segmentation processing based on the first fused data and the second fused data; the model updating module 704 is configured to update the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing, so as to obtain an updated three-dimensional model of the surface of the strip mine.
Illustratively, the surface three-dimensional model construction device of the strip mine further comprises: an obtaining module 705, configured to obtain local point cloud data collected by a collection device for a target area, and perform data preprocessing on the local point cloud data, where the data preprocessing includes: point cloud grouping, time sequence serialization and data deletion processing; accordingly, the first fusion processing module 701 performs the first fusion processing based on the local point cloud data after the data preprocessing, and the second fusion processing module 702 performs the second fusion processing based on the local point cloud data after the data preprocessing.
Exemplarily, the first fusion processing module 701 is specifically configured to:
acquiring an attitude transformation matrix between adjacent frame point clouds;
and performing interframe matching on the local point cloud data based on the attitude transformation matrix to obtain the data after the first fusion processing.
Illustratively, the first fusion processing module 701 is further configured to:
determining the attitude transformation matrix for point cloud data at different moments in the local point cloud data based on linear interpolation transformation; alternatively, the first and second electrodes may be,
the attitude transformation matrix is determined based on detection data of an inertial measurement unit.
Exemplarily, the second fusion processing module 702 is specifically configured to:
carrying out rough matching on the point cloud data in the local point cloud data based on the positioning information of the acquisition equipment;
and carrying out spatial fusion on the roughly matched local point cloud data based on feature identification to obtain data after second fusion processing.
Illustratively, the segmentation processing module 703 is specifically configured to:
carrying out segmentation processing on the geographic object and the non-geographic object on the data subjected to the first fusion processing and the data subjected to the second fusion processing to obtain data from which the non-geographic object is deleted;
wherein the geographic object comprises at least one of: a slope surface and a step surface; the non-geographic object includes at least one of: transport trucks, excavators and bulldozers.
Illustratively, model update module 704 is specifically configured to:
and fusing the data after the non-geographic object is deleted with global point cloud data corresponding to the historical three-dimensional model of the surface of the strip mine, and determining an updated three-dimensional model based on the fused new global point cloud data.
In practical application, the first fusion processing module 701, the second fusion processing module 702, the segmentation processing module 703, the model updating module 704 and the obtaining module 705 can be implemented by a processor in a three-dimensional model building device for the surface of a strip mine. Of course, the processor needs to run a computer program in memory to implement its functions.
It should be noted that: the three-dimensional model building device for surface of strip mine provided in the above embodiments is only illustrated by dividing the above program modules when building the three-dimensional model of surface of strip mine, and in practical applications, the above processes may be distributed to different program modules as needed, that is, the internal structure of the device may be divided into different program modules to complete all or part of the above processes. In addition, the device for building the three-dimensional model of the surface of the strip mine provided by the embodiment and the method for building the three-dimensional model of the surface of the strip mine belong to the same concept, and the specific implementation process is detailed in the method embodiment and is not repeated herein.
Based on the hardware implementation of the program module, in order to implement the method of the embodiment of the invention, the embodiment of the invention also provides surface three-dimensional model building equipment for the strip mine. Fig. 8 shows only an exemplary structure of the apparatus and not the entire structure, and a part of or the entire structure shown in fig. 8 may be implemented as necessary.
As shown in fig. 8, the apparatus 800 for building a three-dimensional model of a surface of a strip mine according to an embodiment of the present invention includes: at least one processor 801, memory 802, a user interface 803, and at least one network interface 804. The various components of the surface three-dimensional model building apparatus 800 are coupled together by a bus system 805. It will be appreciated that the bus system 805 is used to enable communications among the components of the connection. The bus system 805 includes a power bus, a control bus, and a status signal bus in addition to a data bus. For clarity of illustration, however, the various buses are labeled as bus system 805 in fig. 8.
The user interface 803 may include, among other things, a display, a keyboard, a mouse, a trackball, a click wheel, a key, a button, a touch pad, or a touch screen.
The memory 802 in embodiments of the present invention is used to store various types of data to support the operation of the surface three-dimensional model building apparatus. Examples of such data include: any computer program for operating on a three-dimensional model building facility of the surface of a surface mine.
The method for constructing the surface three-dimensional model of the strip mine disclosed by the embodiment of the invention can be applied to the processor 801 or realized by the processor 801. The processor 801 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the method for constructing a three-dimensional model of the surface of a strip mine may be implemented by instructions in the form of hardware integrated logic circuits or software in the processor 801. The Processor 801 may be a general purpose Processor, a Digital Signal Processor (DSP), or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, or the like. Processor 801 may implement or perform the methods, steps, and logic blocks disclosed in embodiments of the present invention. A general purpose processor may be a microprocessor or any conventional processor or the like. The steps of the method disclosed by the embodiment of the invention can be directly implemented by a hardware decoding processor, or can be implemented by combining hardware and software modules in the decoding processor. The software modules may be located in a storage medium located in the memory 802, and the processor 801 reads the information in the memory 802, and completes the steps of the method for building a three-dimensional model of a surface of a strip mine according to the embodiment of the present invention in combination with hardware thereof.
In an exemplary embodiment, the surface three-dimensional model building Device may be implemented by one or more Application Specific Integrated Circuits (ASICs), DSPs, Programmable Logic Devices (PLDs), Complex Programmable Logic Devices (CPLDs), FPGAs, general purpose processors, controllers, Micro Controllers (MCUs), microprocessors (microprocessors), or other electronic components for performing the aforementioned methods.
It will be appreciated that the memory 802 can be either volatile memory or nonvolatile memory, and can include both volatile and nonvolatile memory. Among them, the nonvolatile Memory may be a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), a magnetic random access Memory (FRAM), a Flash Memory (Flash Memory), a magnetic surface Memory, an optical disk, or a Compact Disc Read-Only Memory (CD-ROM); the magnetic surface storage may be disk storage or tape storage. Volatile Memory can be Random Access Memory (RAM), which acts as external cache Memory. By way of illustration and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), Enhanced Synchronous Dynamic Random Access Memory (Enhanced DRAM), Synchronous Dynamic Random Access Memory (SLDRAM), Direct Memory (DRmb Access), and Random Access Memory (DRAM). The described memory for embodiments of the present invention is intended to comprise, without being limited to, these and any other suitable types of memory.
In order to realize the method of the embodiment of the invention, the embodiment of the invention also provides a surface three-dimensional model building system of the strip mine in the strip scene. Fig. 9 illustrates only an exemplary structure of the system, not the entire structure, and a part or the entire structure illustrated in fig. 9 may be implemented as necessary.
As shown in fig. 9, a 9-strip mine surface three-dimensional model building system 900 according to an embodiment of the present invention includes: a radar loading device 901, a mine management scheduling center 902, and a cloud computing center 903 (equivalent to the three-dimensional model building device 800 described above). The components of the radar loading device 901, the mine management dispatching center 902 and the cloud computing center 903 in the surface three-dimensional model building system 900 of the strip mine are coupled together through a bus, and the bus is also used for realizing the connection and communication among the components. The bus includes a power bus, a control bus, and a status signal bus in addition to a data bus. A5G base station or a Wi-Fi9 local area network is established in a mine, the average transmission rate can reach 10Gbps, point cloud data sent back by hundreds of devices can be received simultaneously, and the data transmission requirement of the three-dimensional automatic modeling technology of the medium-surface mining scene can be met.
The radar loading apparatus 901 includes: a data acquisition interface 9011, a first data transceiving interface 9012, and a user interface 9013. The data acquisition interface 9011 is configured to collect, through each sensor or scanning device, data information corresponding to an outdoor scene in the embodiment of the present invention. The first data transceiving interface 9012 is in communication interaction with the mine management dispatching center 902 through a bus. The first data transceiving interface 9012 may include a communication antenna. The user interface 9013 may include a display, a keyboard, a mouse, a trackball, a click wheel, keys, buttons, a touch pad, a touch screen, or the like, for enabling user interaction with the system.
The mine management scheduling center 902 includes: a second data transceiving interface 9021 and a first data processing interface 9022. The second data transceiving interface 9022 transmits data information to the cloud computing center 903 and receives data information from the first data transceiving interface 9012 through the bus. The first data processing interface 9022 is configured to process data information from the radar loading device 901 and the cloud computing center 903, and perform error analysis and preprocessing of confidence calculation on data collected by a sensor installed on the surface mine device. The first data processing interface 9022 includes a processor and a storage medium storing a computer program, wherein the processor executes the computer program stored in the storage medium to perform preprocessing on data collected by the sensor.
The cloud computing center 903 includes: a third data transceiving interface 9031 and a second data processing interface 9032. The third data transceiving interface 9031 receives data information sent from the second data transceiving interface 9021 through the bus, and transmits the processed result and instruction to the second data transceiving interface 9021. The second data processing interface 9032 receives data information from the third data transceiving interface 9031, and generates a three-dimensional model of the strip mine with seamless geometry and separated attributes by running programs such as point cloud fusion, non-geographic object separation, three-dimensional modeling and the like in the functional area. The second data processing interface 9032 includes a mass processor and a storage medium storing a computer program, the processor running the computer program in the storage medium to implement three-dimensional automatic modeling of a surface mining scenario.
In an exemplary embodiment, the embodiment of the present invention further provides a storage medium, specifically a computer storage medium, which may be a computer readable storage medium, for example, a memory 802 storing a computer program, which is executable by a processor 801 of a surface three-dimensional model building apparatus of a strip mine to perform the steps of the method according to the embodiment of the present invention. The computer readable storage medium may be a ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface Memory, optical disk, or CD-ROM, among others.
It should be noted that: "first," "second," and the like are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.
In addition, the technical solutions described in the embodiments of the present invention may be arbitrarily combined without conflict.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A method for constructing a three-dimensional model of the surface of a strip mine is characterized by comprising the following steps:
performing first fusion processing on local point cloud data acquired by acquisition equipment aiming at a target area based on a time dimension;
performing second fusion processing on the local point cloud data based on the spatial dimension;
performing target segmentation processing based on the data after the first fusion processing and the data after the second fusion processing;
and updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine.
2. The method of claim 1, wherein prior to the first fusing process and the second fusing process, the method further comprises:
acquiring local point cloud data acquired by acquisition equipment aiming at a target area;
performing data preprocessing on the local point cloud data, wherein the data preprocessing comprises the following steps: point cloud grouping, time sequence serialization and data deletion processing;
correspondingly, the first fusion processing and the second fusion processing are carried out on the basis of the local point cloud data after the data preprocessing.
3. The method of claim 1, wherein the performing a first fusion process on the local point cloud data acquired by the acquisition device for the target area based on the time dimension comprises:
acquiring an attitude transformation matrix between adjacent frame point clouds;
and performing interframe matching on the local point cloud data based on the attitude transformation matrix to obtain the data after the first fusion processing.
4. The method of claim 3, further comprising:
determining the attitude transformation matrix for point cloud data at different moments in the local point cloud data based on linear interpolation transformation; alternatively, the first and second electrodes may be,
the attitude transformation matrix is determined based on detection data of an inertial measurement unit.
5. The method of claim 1, wherein the second fusing of the local point cloud data based on spatial dimensions comprises:
carrying out rough matching on the point cloud data in the local point cloud data based on the positioning information of the acquisition equipment;
and carrying out spatial fusion on the roughly matched local point cloud data based on feature identification to obtain data after second fusion processing.
6. The method according to claim 1, wherein performing a target segmentation process based on the first fusion processed data and the second fusion processed data comprises:
carrying out segmentation processing on the geographic object and the non-geographic object on the data subjected to the first fusion processing and the data subjected to the second fusion processing to obtain data from which the non-geographic object is deleted;
wherein the geographic object comprises at least one of: a slope surface and a step surface; the non-geographic object includes at least one of: transport trucks, excavators and bulldozers.
7. The method of claim 6, wherein updating the historical three-dimensional model of the surface mine based on the target segmentation processed data to obtain an updated three-dimensional model of the surface mine comprises:
and fusing the data after the non-geographic object is deleted with global point cloud data corresponding to the historical three-dimensional model of the surface of the strip mine, and determining an updated three-dimensional model based on the fused new global point cloud data.
8. A three-dimensional model building device for surface of strip mine is characterized by comprising the following components:
the first fusion processing module is used for performing first fusion processing on local point cloud data acquired by the acquisition equipment aiming at the target area based on the time dimension;
the second fusion processing module is used for carrying out second fusion processing on the local point cloud data based on the space dimension;
the segmentation processing module is used for carrying out target segmentation processing on the basis of the data subjected to the first fusion processing and the data subjected to the second fusion processing;
and the model updating module is used for updating the historical three-dimensional model of the surface of the strip mine based on the data after the target segmentation processing to obtain the updated three-dimensional model of the surface of the strip mine.
9. An open pit mine earth surface three-dimensional model construction device is characterized by comprising: a processor and a memory for storing a computer program capable of running on the processor, wherein,
the processor, when executing the computer program, is adapted to perform the steps of the method of any of claims 1 to 7.
10. A storage medium having a computer program stored thereon, the computer program, when executed by a processor, implementing the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110879529.5A CN113593017A (en) | 2021-08-02 | 2021-08-02 | Method, device and equipment for constructing surface three-dimensional model of strip mine and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110879529.5A CN113593017A (en) | 2021-08-02 | 2021-08-02 | Method, device and equipment for constructing surface three-dimensional model of strip mine and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113593017A true CN113593017A (en) | 2021-11-02 |
Family
ID=78253571
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110879529.5A Pending CN113593017A (en) | 2021-08-02 | 2021-08-02 | Method, device and equipment for constructing surface three-dimensional model of strip mine and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113593017A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114067073A (en) * | 2021-11-16 | 2022-02-18 | 安徽理工大学 | TLS point cloud-based mining area building deformation automatic extraction method |
| CN114584571A (en) * | 2021-12-24 | 2022-06-03 | 北京中电飞华通信有限公司 | Power grid station digital twin synchronous communication method based on spatial computing technology |
| CN114609591A (en) * | 2022-03-18 | 2022-06-10 | 湖南星晟智控科技有限公司 | Data processing method based on laser point cloud data |
| CN114913352A (en) * | 2022-05-05 | 2022-08-16 | 山东高速建设管理集团有限公司 | Multi-source information space-time registration method and system based on joint similarity matching |
| CN115375866A (en) * | 2022-10-20 | 2022-11-22 | 国家电投集团科学技术研究院有限公司 | Method, device, equipment and medium for updating three-dimensional geological model of mining area |
| CN115409941A (en) * | 2022-08-31 | 2022-11-29 | 中南大学 | Three-dimensional ground object model fusion method and system in three-dimensional road scene |
| CN115509406A (en) * | 2022-11-23 | 2022-12-23 | 煤炭科学研究总院有限公司 | Coal mine multi-scene fusion method and device, storage medium and electronic equipment |
| CN116067379A (en) * | 2023-03-07 | 2023-05-05 | 青岛慧拓智能机器有限公司 | Multi-sensor fusion positioning method based on dynamic point cloud under long tunnel scene |
-
2021
- 2021-08-02 CN CN202110879529.5A patent/CN113593017A/en active Pending
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114067073B (en) * | 2021-11-16 | 2022-09-06 | 安徽理工大学 | TLS point cloud-based mining area building deformation automatic extraction method |
| CN114067073A (en) * | 2021-11-16 | 2022-02-18 | 安徽理工大学 | TLS point cloud-based mining area building deformation automatic extraction method |
| CN114584571A (en) * | 2021-12-24 | 2022-06-03 | 北京中电飞华通信有限公司 | Power grid station digital twin synchronous communication method based on spatial computing technology |
| CN114584571B (en) * | 2021-12-24 | 2024-02-27 | 北京中电飞华通信有限公司 | Space calculation technology-based digital twin synchronous communication method for power grid station |
| CN114609591A (en) * | 2022-03-18 | 2022-06-10 | 湖南星晟智控科技有限公司 | Data processing method based on laser point cloud data |
| CN114609591B (en) * | 2022-03-18 | 2022-12-20 | 湖南星晟智控科技有限公司 | Data processing method based on laser point cloud data |
| CN114913352B (en) * | 2022-05-05 | 2023-05-26 | 山东高速建设管理集团有限公司 | Multi-source information space-time registration method and system based on joint similarity matching |
| CN114913352A (en) * | 2022-05-05 | 2022-08-16 | 山东高速建设管理集团有限公司 | Multi-source information space-time registration method and system based on joint similarity matching |
| CN115409941A (en) * | 2022-08-31 | 2022-11-29 | 中南大学 | Three-dimensional ground object model fusion method and system in three-dimensional road scene |
| CN115375866B (en) * | 2022-10-20 | 2023-02-03 | 国家电投集团科学技术研究院有限公司 | Method, device, equipment and medium for updating three-dimensional geological model of mining area |
| CN115375866A (en) * | 2022-10-20 | 2022-11-22 | 国家电投集团科学技术研究院有限公司 | Method, device, equipment and medium for updating three-dimensional geological model of mining area |
| CN115509406B (en) * | 2022-11-23 | 2023-03-14 | 煤炭科学研究总院有限公司 | Coal mine multi-scene fusion method and device, storage medium and electronic equipment |
| CN115509406A (en) * | 2022-11-23 | 2022-12-23 | 煤炭科学研究总院有限公司 | Coal mine multi-scene fusion method and device, storage medium and electronic equipment |
| CN116067379A (en) * | 2023-03-07 | 2023-05-05 | 青岛慧拓智能机器有限公司 | Multi-sensor fusion positioning method based on dynamic point cloud under long tunnel scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113593017A (en) | Method, device and equipment for constructing surface three-dimensional model of strip mine and storage medium | |
| CN108763287B (en) | Construction method of large-scale passable regional driving map and unmanned application method thereof | |
| Mühlfellner et al. | Summary maps for lifelong visual localization | |
| CN113168717B (en) | Point cloud matching method and device, navigation method and equipment, positioning method and laser radar | |
| Majdik et al. | Air‐ground matching: Appearance‐based GPS‐denied urban localization of micro aerial vehicles | |
| US8437501B1 (en) | Using image and laser constraints to obtain consistent and improved pose estimates in vehicle pose databases | |
| CN112380312B (en) | Laser map updating method based on grid detection, terminal and computer equipment | |
| KR102200299B1 (en) | A system implementing management solution of road facility based on 3D-VR multi-sensor system and a method thereof | |
| CN114842438A (en) | Terrain detection method, system and readable storage medium for autonomous driving vehicle | |
| CN111710040B (en) | High-precision map construction method, system, terminal and storage medium | |
| US11861855B2 (en) | System and method for aerial to ground registration | |
| CN109871739B (en) | Automatic target detection and space positioning method for mobile station based on YOLO-SIOCTL | |
| CN113516664A (en) | Visual SLAM method based on semantic segmentation dynamic points | |
| CN115294293B (en) | Method for automatically compiling high-precision map road reference line based on low-altitude aerial photography result | |
| Sun et al. | Roads and intersections extraction from high-resolution remote sensing imagery based on tensor voting under big data environment | |
| Guo et al. | Extraction of dense urban buildings from photogrammetric and LiDAR point clouds | |
| CN114022760B (en) | Railway tunnel barrier monitoring and early warning method, system, equipment and storage medium | |
| CN115638787A (en) | Digital map generation method, computer readable storage medium and electronic device | |
| CN113096181A (en) | Method and device for determining pose of equipment, storage medium and electronic device | |
| JP2020153956A (en) | Mobile location estimation system and mobile location method | |
| CN111612829B (en) | High-precision map construction method, system, terminal and storage medium | |
| CN114115231B (en) | Space pose point cloud correction method and system for mobile robot | |
| CN113763468B (en) | Positioning method, device, system and storage medium | |
| CN113227713A (en) | Method and system for generating environment model for positioning | |
| CN115147549A (en) | Urban three-dimensional model generation and updating method based on multi-source data fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |