CN-121978713-A - Terrain mapping data processing method for unmanned aerial vehicle-mounted LiDAR

CN121978713ACN 121978713 ACN121978713 ACN 121978713ACN-121978713-A

Abstract

The invention relates to the technical field of surveying and mapping remote sensing, and provides a terrain surveying and mapping data processing method of an unmanned aerial vehicle-mounted LiDAR, which comprises the steps of receiving original point cloud data and inertial measurement data, and performing attitude accurate compensation on the inertial measurement data to obtain inertial navigation error compensation parameters; the method comprises the steps of performing kinematic trajectory reconstruction and trajectory constraint optimization on original point cloud data to obtain high-precision point cloud data, performing multi-echo characteristic analysis and penetrability filtering on the high-precision point cloud data to obtain pure earth surface point cloud data, performing terrain gradient classification on a mapping area to obtain a horizontal terrain area and an inclined terrain area, performing equidistant thinning on the horizontal terrain area to obtain thinning point cloud data, performing characteristic line encryption on the inclined terrain area to obtain encrypted point cloud data, and performing spatial data mosaic on the thinning point cloud data and the encrypted point cloud data to obtain three-dimensional digital terrain data.

Inventors

CHEN DONGJIE
XIANG GUO
JIANG ZHENPENG
HU ZHENLUAN
Jia Xianjuan
ZHU ZITONG

Assignees

淄博金宅测绘有限公司

Dates

Publication Date: 20260505
Application Date: 20260324

Claims (10)

1. A method for processing topographic mapping data of an unmanned aerial vehicle-mounted LiDAR, the method comprising: p1, receiving original point cloud data acquired by an unmanned aerial vehicle airborne radar in a mapping area, synchronously acquiring inertial measurement data of the unmanned aerial vehicle, and performing attitude accurate compensation on the inertial measurement data to obtain inertial navigation error compensation parameters of the unmanned aerial vehicle; P2, performing kinematic trajectory reconstruction on positioning information in the original point cloud data based on the inertial navigation error compensation parameters, and performing trajectory constraint optimization on the original point cloud data according to the reconstructed positioning information to obtain high-precision point cloud data of the mapping region; p3, performing multi-echo characteristic analysis on the high-precision point cloud data, and performing penetrability filtering on the analyzed primary vegetation point cloud clusters to obtain pure earth surface point cloud data of the mapping area; p4, classifying the terrain gradient of the mapping area based on the pure surface point cloud data to obtain a horizontal terrain area and an inclined terrain area of the mapping area; p5, performing equidistant thinning on the point cloud in the horizontal terrain area to obtain thinned point cloud data of the horizontal terrain area; And P6, carrying out characteristic line encryption on the point cloud in the inclined terrain area to obtain encrypted point cloud data of the inclined terrain area, and carrying out space data mosaic on the sparse point cloud data and the encrypted point cloud data to obtain three-dimensional digital terrain data of the mapping area.
2. The method for processing topographic mapping data of LiDAR carried by an unmanned aerial vehicle according to claim 1, wherein the steps of receiving original point cloud data collected by the unmanned aerial vehicle carried radar in a mapping area, synchronously acquiring inertial measurement data of the unmanned aerial vehicle, and carrying out attitude accurate compensation on the inertial measurement data to obtain inertial navigation error compensation parameters of the unmanned aerial vehicle comprise: receiving original point cloud data acquired by an unmanned aerial vehicle airborne radar in a mapping area, wherein the original point cloud data comprises three-dimensional space coordinates, laser echo intensity values and multi-echo frequency identifications; acquiring inertial measurement data of the unmanned aerial vehicle, wherein the inertial measurement data comprises a triaxial angular velocity output sequence, a triaxial acceleration output sequence and an inertial navigation system time reference mark; Performing time base line registration on the triaxial angular velocity output sequence and the triaxial acceleration output sequence to obtain time base registration inertial navigation data of the unmanned aerial vehicle; Based on the time base registration inertial navigation data, extracting the static drift amount of the unmanned aerial vehicle in a hovering stage, and performing zero offset trend fitting on the static drift amount to obtain a gyroscope zero offset trend line of the unmanned aerial vehicle; Performing differential offset on the zero offset trend line of the gyroscope and the angular velocity component in the time-base registration inertial navigation data, and performing angle increment accumulation on the offset angular velocity data to obtain a preliminary attitude angle sequence of the unmanned aerial vehicle; And performing cone error compensation on the preliminary attitude angle sequence to obtain inertial navigation error compensation parameters of the unmanned aerial vehicle.
3. The method for processing topographic mapping data of an unmanned aerial vehicle on-board LiDAR according to claim 2, wherein the steps of differentially canceling angular velocity components in the gyroscope zero offset trend line and the time-based registration inertial navigation data, and performing angular increment accumulation on the angular velocity data after cancellation to obtain a preliminary attitude angle sequence of the unmanned aerial vehicle comprise the steps of: Extracting an original angular velocity instantaneous value of an angular velocity component in the time-base registration inertial navigation data, and carrying out zero offset compensation on the original angular velocity instantaneous value according to the gyroscope zero offset trend line to obtain a pure angular velocity value of the unmanned aerial vehicle; performing angular velocity accumulation on the pure angular velocity value to obtain an angular increment value of the unmanned aerial vehicle; and updating the quaternion of the angle increment value to obtain a gesture quaternion of the unmanned aerial vehicle, wherein the calculation formula of the gesture quaternion is as follows: ; Wherein, the Represent the first The gesture quaternions for the individual sampling instants, Represent the first The gesture quaternions for the individual sampling instants, Representation of Sampling time and the first The angular increment value between sampling instants, Representation of Is used for the mold length of the mold, Representing a quaternion multiplication; and carrying out Euler angle conversion on the attitude quaternion to obtain a preliminary attitude angle sequence of the unmanned aerial vehicle.
4. The method for processing topographic mapping data of an unmanned aerial vehicle-mounted LiDAR according to claim 2, wherein the performing kinematic trajectory reconstruction on the positioning information in the raw point cloud data based on the inertial navigation error compensation parameter, and performing trajectory constraint optimization on the raw point cloud data according to the reconstructed positioning information, to obtain high-precision point cloud data of the mapping region comprises: Performing point cloud frame interpretation on the original point cloud data to obtain intra-frame laser points of the mapping area, wherein the intra-frame laser points carry the three-dimensional space coordinates and the corresponding time labels; Based on the inertial navigation error compensation parameter, performing motion distortion correction on the three-dimensional space coordinate to obtain an instantaneous positioning coordinate of the laser point in the frame; performing track recursion on the instantaneous positioning coordinates based on the sequence of the corresponding time labels to obtain an original flight band of the unmanned aerial vehicle; Performing closed loop annular space deviation checking on the original flight zone, and performing overall constraint adjustment on the original flight zone according to a checking result to obtain an optimized flight zone of the unmanned aerial vehicle; And re-projecting the laser points to the space positions corresponding to the optimized flight zone based on the optimized flight zone so as to obtain high-precision point cloud data of the mapping area.
5. The method for processing topographic mapping data of LiDAR carried by an unmanned aerial vehicle according to claim 4, wherein the steps of performing closed loop annular time deviation checking on the original flight zone, and performing overall constraint adjustment on the original flight zone according to a checking result to obtain an optimized flight zone of the unmanned aerial vehicle comprise: extracting an overlapping region between adjacent bands in the original flight band; Based on the overlapping region, performing space pose registration on the instantaneous positioning coordinates to obtain a space deviation vector of the overlapping region; According to the space deviation vector, performing topology constraint networking on the navigation belt nodes of the original flight navigation belt to obtain a navigation belt network topology relation diagram of the navigation belt nodes; performing global consistency traversal on the navigation belt network topological relation diagram, and constructing a node deviation tensor field of the original flight navigation belt according to the traversed navigation belt nodes; and based on the node deviation tensor field, performing nonlinear relaxation adjustment on the original flight band to obtain the optimized flight band of the unmanned aerial vehicle.
6. The method for processing topographic mapping data of LiDAR on board an unmanned aerial vehicle according to claim 2, wherein the steps of performing multi-echo characteristic analysis on the high-precision point cloud data, and performing penetration filtering on the analyzed primary vegetation point cloud clusters to obtain pure surface point cloud data of the mapping region comprise: based on the multi-echo frequency identification and the laser echo intensity value, carrying out echo attribute clustering on data points in the high-precision point cloud data to obtain a first echo point cloud cluster and a last echo point cloud cluster of the mapping area; Performing echo correlation reconstruction on the first echo point cloud cluster and the last echo point cloud cluster to obtain a primary vegetation point cloud cluster of the mapping area; and carrying out layered penetration filtering on the preliminary vegetation point cloud cluster to obtain pure earth surface point cloud data of the mapping area.
7. The method for processing the topographic survey and drawing data of the LiDAR on board the unmanned aerial vehicle according to claim 2, wherein the classifying the topographic gradient of the survey and drawing area based on the pure surface point cloud data to obtain the horizontal topographic area and the inclined topographic area of the survey and drawing area comprises: based on the three-dimensional space coordinates, carrying out neighborhood gradient field analysis on the surface data points in the pure surface point cloud data to obtain local gradient field values of the surface data points; performing terrain self-adaptive segmentation on the surface data points by taking the local gradient field values as classification primitives to obtain horizontal terrain point cloud clusters and inclined terrain point cloud clusters of the mapping region; And marking the space areas covered by the horizontal terrain point cloud cluster and the inclined terrain point cloud cluster as a horizontal terrain area and an inclined terrain area of the mapping area respectively.
8. The method for processing topographic mapping data of an airborne LiDAR of an unmanned aerial vehicle according to claim 2, wherein the performing equidistant thinning on the point cloud in the horizontal topographic region to obtain the thinned point cloud data of the horizontal topographic region comprises: performing density field interpolation construction on the horizontal terrain area based on the three-dimensional space coordinates to obtain a space distribution density field of the horizontal terrain area; Taking the spatial distribution density field as a weight field, and uniformly sampling the horizontal terrain area by weighting to obtain an initial sampling point cloud of the horizontal terrain area; Performing adjacency relation matching on the spatial adjacent points of the initial sampling point cloud to construct a spatial adjacency graph of the horizontal terrain area; and carrying out space redundancy reduction on the space adjacent graph to obtain the sparse point cloud data of the horizontal terrain area.
9. The method for processing topographic mapping data of an unmanned aerial vehicle-mounted LiDAR according to claim 1, wherein the performing feature line encryption on the point cloud in the inclined topographic region to obtain encrypted point cloud data of the inclined topographic region, and performing spatial data mosaic on the thinned point cloud data and the encrypted point cloud data to obtain three-dimensional digital topographic data of the mapping region comprises: Performing geometric feature deconstructment on the inclined terrain point cloud clusters in the inclined terrain area to obtain a terrain feature skeleton of the inclined terrain area; Based on the terrain feature skeleton, carrying out normal constraint densification on the inclined terrain point cloud cluster to obtain encrypted point cloud data of the inclined terrain area; Performing feature constraint matching on the sparse point cloud data and the encrypted point cloud data to obtain a spatial registration constraint pair of the sparse point cloud data and the encrypted point cloud data; based on the spatial registration constraint pair, performing rigid transformation registration on the sparse point cloud data and the encrypted point cloud data to obtain aligned point cloud data of the sparse point cloud data and the encrypted point cloud data; And performing space mosaic fusion on the alignment point cloud data to obtain three-dimensional digital topographic data of the mapping area.
10. The method for processing topographic mapping data of an unmanned aerial vehicle-mounted LiDAR of claim 9, wherein the performing normal constraint densification on the oblique topographic point cloud cluster based on the topographic feature skeleton to obtain the encrypted point cloud data of the oblique topographic region comprises: Performing normal fitting estimation on discrete skeleton points in the terrain feature skeleton to obtain normal vectors of the discrete skeleton points; carrying out neighborhood expansion on the two sides along the normal vector by taking the discrete skeleton point as a center to obtain a normal neighborhood point of the discrete skeleton point; performing density compensation interpolation on the normal neighborhood points to obtain refined sampling encryption points of the normal neighborhood points; and carrying out space field fusion on the fine sampling encryption points to obtain encryption point cloud data of the inclined terrain area.

Description

Terrain mapping data processing method for unmanned aerial vehicle-mounted LiDAR Technical Field The invention relates to the technical field of surveying and mapping remote sensing, in particular to a terrain surveying and mapping data processing method of an unmanned aerial vehicle-mounted LiDAR. Background In the field of unmanned aerial vehicle-mounted LiDAR topographic mapping data processing, in the prior art, a fine processing mechanism is lacking in attitude compensation of unmanned aerial vehicle inertial measurement data, accurate fitting and error offset are not carried out on static drift amount of an inertial navigation system, special cone error compensation is not carried out on an attitude angle sequence, so that accuracy of inertial navigation error compensation parameters is insufficient, point cloud positioning information track reconstruction and constraint optimization based on the parameters are difficult to eliminate motion distortion of original point cloud, precision of navigation belt adjustment and point cloud projection is limited, high-precision point cloud data of a mapping area cannot be effectively obtained, meanwhile, systematicness is lacking in clustering and association reconstruction design of multi-echo point cloud, penetrability filtering effect of vegetation point cloud is poor, and vegetation interference is difficult to be accurately removed to obtain pure earth surface point cloud data. The prior art only depends on single threshold value judgment for classification of the terrain gradient, lacks self-adaptive segmentation means based on actual characteristics of the terrain, adopts unified processing strategies for point clouds of horizontal and inclined terrains, does not develop targeted thinning and encryption operations according to the terrain characteristics, is easy to cause redundancy of horizontal terrain point cloud data, and can cause loss of inclined terrain characteristic point clouds, so that finally generated three-dimensional digital terrain data cannot accurately restore the topography and original appearance of a mapping area, and meanwhile, more redundant operations exist in an overall data processing link, so that the processing efficiency of unmanned aerial vehicle-mounted LiDAR topographic mapping data is at a lower level, and therefore, how to improve the accuracy and efficiency of unmanned aerial vehicle-mounted LiDAR topographic mapping data processing becomes a problem to be solved urgently. Disclosure of Invention The invention provides a topographic mapping data processing method of an unmanned aerial vehicle-mounted LiDAR, which aims to solve the problems in the background technology. In order to achieve the above object, the present invention provides a method for processing topographic mapping data of an unmanned aerial vehicle-mounted LiDAR, comprising: p1, receiving original point cloud data acquired by an unmanned aerial vehicle airborne radar in a mapping area, synchronously acquiring inertial measurement data of the unmanned aerial vehicle, and performing attitude accurate compensation on the inertial measurement data to obtain inertial navigation error compensation parameters of the unmanned aerial vehicle; P2, performing kinematic trajectory reconstruction on positioning information in the original point cloud data based on the inertial navigation error compensation parameters, and performing trajectory constraint optimization on the original point cloud data according to the reconstructed positioning information to obtain high-precision point cloud data of the mapping region; p3, performing multi-echo characteristic analysis on the high-precision point cloud data, and performing penetrability filtering on the analyzed primary vegetation point cloud clusters to obtain pure earth surface point cloud data of the mapping area; p4, classifying the terrain gradient of the mapping area based on the pure surface point cloud data to obtain a horizontal terrain area and an inclined terrain area of the mapping area; p5, performing equidistant thinning on the point cloud in the horizontal terrain area to obtain thinned point cloud data of the horizontal terrain area; And P6, carrying out characteristic line encryption on the point cloud in the inclined terrain area to obtain encrypted point cloud data of the inclined terrain area, and carrying out space data mosaic on the sparse point cloud data and the encrypted point cloud data to obtain three-dimensional digital terrain data of the mapping area. In a preferred embodiment, the receiving the raw point cloud data collected by the unmanned aerial vehicle airborne radar in the mapping area, synchronously obtaining the inertial measurement data of the unmanned aerial vehicle, and performing attitude accurate compensation on the inertial measurement data to obtain the inertial navigation error compensation parameter of the unmanned aerial vehicle, including: receiving original point cloud da