CN115018973A - Low-altitude unmanned-machine point cloud modeling precision target-free evaluation method - Google Patents
Low-altitude unmanned-machine point cloud modeling precision target-free evaluation method Download PDFInfo
- Publication number
- CN115018973A CN115018973A CN202210387155.XA CN202210387155A CN115018973A CN 115018973 A CN115018973 A CN 115018973A CN 202210387155 A CN202210387155 A CN 202210387155A CN 115018973 A CN115018973 A CN 115018973A
- Authority
- CN
- China
- Prior art keywords
- point cloud
- point
- aerial vehicle
- unmanned aerial
- elevation
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/003—Navigation within 3D models or images
Abstract
The invention discloses a target-free evaluation method for low-altitude unmanned aerial vehicle point cloud modeling precision, which comprises the steps of collecting unmanned aerial vehicle laser point cloud data and measuring a control point; acquiring a characteristic point GPS coordinate through two-dimensional modeling of unmanned aerial vehicle laser point cloud data; acquiring a feature point elevation through three-dimensional modeling of unmanned aerial vehicle laser point cloud data; and comparing the GPS coordinates of the characteristic points with the GPS coordinates of the control points, and comparing the elevation of the characteristic points with the elevation of the control points to finish the target-free evaluation. According to the invention, the target does not need to be laid during the operation of the unmanned aerial vehicle, the subsequent point cloud modeling precision evaluation can be completed, the time and labor for manually carrying the target in the field are avoided, and great convenience is brought to the operation of the unmanned aerial vehicle in the field.
Description
Technical Field
The invention relates to the technical field of aerial remote sensing measurement, in particular to a target-free evaluation method for low-altitude unmanned point cloud modeling precision.
Background
The area generally applied by the traditional aerial photogrammetry is a larger range, and for a smaller area, the aerial photogrammetry technology depends on an airport and weather conditions to a great extent, and the expense is larger, the aerial photography period is longer, and the aerial photogrammetry technology is not beneficial to playing the role of the aerial photogrammetry technology. The technology of using the unmanned aerial vehicle as an aerial remote sensing platform can fill up the defects of the traditional aerial photogrammetry.
The unmanned aerial vehicle can flexibly and inexpensively acquire high-image-resolution images in a tilt photogrammetry mode. Meanwhile, the mode that the unmanned aerial vehicle carries the attitude determination and orientation system and the laser radar system is adopted to carry out rapid and accurate measurement operation in a small area, and the method has great advantages. The laser radar emits laser beams to perform active scanning, so that complex scenes can be comprehensively perceived, and the complex scenes can be presented and restored in a high-definition and high-precision mode. The point cloud data reflects the information such as appearance characteristics, height and position of a research object with real effect and surveying and mapping level precision, can further generate products with more added values such as DEM (digital elevation model) and DTM (data transformation model), and plays an important role in the fields of surveying and mapping geographic information, disaster prevention and reduction, agricultural estimation, water conservancy and power, traffic and the like.
However, in the field unmanned aerial vehicle operation, in order to verify the accuracy of data acquisition, a target is often laid on the ground, so that the position of a control point can be confirmed conveniently in the subsequent modeling process. In rugged mountain environments or extremely cold polar regions, the operation of carrying the target is time-consuming and labor-consuming, loose soil may occur in some control point regions, and no foot drop points exist, which brings great challenges to the laying of the target. Therefore, the invention provides a target-free evaluation method for low-altitude unmanned aerial vehicle point cloud modeling precision, and the method can finish the evaluation of the subsequent point cloud modeling precision without laying a target during the operation of an unmanned aerial vehicle, thereby bringing great convenience to the operation of a field unmanned aerial vehicle.
Disclosure of Invention
This section is for the purpose of summarizing some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. In this section, as well as in the abstract and the title of the invention of this application, simplifications or omissions may be made to avoid obscuring the purpose of the section, the abstract and the title, and such simplifications or omissions are not intended to limit the scope of the invention.
The present invention has been made in view of the above and/or other problems occurring in the existing unmanned aerial vehicle photogrammetry method.
Therefore, the invention aims to solve the problem that a target-free evaluation method of low-altitude unmanned aerial vehicle point cloud modeling precision is needed, and the problem that the existing field unmanned aerial vehicle needs to lay a target during operation, so that time and labor are wasted is solved.
In order to solve the technical problems, the invention provides the following technical scheme: a target-free evaluation method for low-altitude unmanned aerial vehicle point cloud modeling precision comprises the steps of collecting unmanned aerial vehicle laser point cloud data and measuring control points; acquiring a characteristic point GPS coordinate through two-dimensional modeling of unmanned aerial vehicle laser point cloud data; acquiring a feature point elevation through three-dimensional modeling of unmanned aerial vehicle laser point cloud data; and comparing the GPS coordinates of the characteristic points with the GPS coordinates of the control points, and comparing the elevation of the characteristic points with the elevation of the control points to finish the target-free evaluation.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: the collection of unmanned aerial vehicle laser point cloud data adopts unmanned aerial vehicle to carry on radar camera lens and gathers, and concrete step is: preparing before flying, setting an area on virtual earth software, storing a KML file, and importing the KML file to obtain a flying designated area; planning air routes, namely determining a measuring area range after considering surrounding environment factors through preliminary design and on-site investigation planning; setting specific parameters according to different target task requirements, wherein the parameters comprise RTK and lens parameters and flight parameters; after the parameters are set, whether the unmanned aerial vehicle is connected with a network RTK or not is confirmed, whether the radar lens completes inertial navigation preheating or not is confirmed, and data collection is carried out after takeoff is confirmed.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: the measurement of the control points comprises selecting the control points, checking the stability and integrity of the existing control points for the existing control points, selecting the ground control points at the positions where the measurement is convenient to store and implement, and widening the view above the control points; and measuring and recording the GPS coordinates and elevation information of the control points by using the GNSS receiver.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: the two-dimensional modeling of the laser point cloud data of the unmanned aerial vehicle comprises the following specific steps: adopting unmanned aerial vehicle mapping software to obtain high-precision DOM data; unifying DOM data and control point coordinates to a WGS 84/UTM Zone 50N coordinate system in a GIS platform; positioning a characteristic point based on DOM data, positioning a characteristic feature visible to the naked eye near a control point based on visual judgment in a geographic information system, and taking two crossed sidelines of the characteristic feature as an extension line, wherein the intersection point of the two extension lines is the characteristic point which is the position of the control point on a DOM image; and calculating the horizontal distance deviation delta x and the vertical distance deviation delta y of the characteristic points and the control point GPS coordinates through pixel level comparison, and finally representing the modeled positioning error by using the straight line distance deviation delta l.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: the three-dimensional modeling of the laser point cloud data of the unmanned aerial vehicle comprises the following specific steps: creating a three-dimensional model through unmanned aerial vehicle mapping software; converting the format of the three-dimensional model; intercepting point clouds in different ranges by taking the characteristic points as circle centers; using chi-square fitness test to remove point clouds with chi-square values larger than a critical value, and reducing gross errors; and performing self-adaptive polynomial fitting on the remaining point cloud to obtain elevation information of the characteristic points, comparing the elevation information with the elevation information of the control points in actual measurement, and representing the modeling elevation error by using delta h.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: the format conversion of the three-dimensional model specifically comprises the following steps: and importing the generated three-dimensional model into point cloud processing software, performing cutting and data conversion on the three-dimensional model, converting the three-dimensional model into a txt file with GPS coordinates and elevation information, and importing the txt file into a GIS platform for point spread.
As an optimal scheme of the low-altitude unmanned point cloud modeling precision target-free evaluation method, the method comprises the following steps: intercepting point clouds in different ranges by taking the characteristic points as circle centers, specifically comprising the following steps:
the method comprises the following steps of searching point clouds near characteristic points by taking the characteristic points as circle centers and different lengths as search radiuses l, and reconstructing a three-dimensional terrain ground by using a self-adaptive polynomial, wherein the self-adaptive polynomial is as follows:
Z＝h c +a 1 x+a 2 y+a 3 x 2 +a 4 y 2 +a 5 xy+a 6 x 3 +a 7 y 3 +a 8 x 2 y+a 9 xy 2
wherein x and y are coordinates of the point cloud under a WGS 84/UTM Zone 50N coordinate system respectively, the order of the adaptive polynomial is determined to be 3, and 10 parameters are used for fitting the curved surface in order to avoid the multi-fitting phenomenon of the curved surface.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: the adaptive polynomial may be converted to solve for 10 parameters as follows:
wherein Z is the elevation of each point cloud, e is the observation error, which is a range between
The value of (c). Sigma 0 The unit weight is P, the weight matrix of all point clouds in the search radius is obtained by giving different weights to each point cloud according to the error value of the elevation of each point cloud and the elevation of the characteristic point;
b is a polynomial combination form of x and y with different orders, and is represented by the following formula:
when the curved surface is fitted, a local coordinate system is established by taking the characteristic point as an origin, and the coordinate of the characteristic point is expressed as (X) c ,Y c ) The coordinates of the other point clouds relative to the coordinates of the feature points are expressed as (X-X) c ,y-Y c ) L is a search radius, and the precision and the efficiency of least square solution are improved by dividing the search radius by a proportionality coefficient (taking a search radius value as reference);
m and n represent orders, and the highest order is not more than 3.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: in the conversion formula, X is a parametric equation to be solved, and is represented by the following formula:
wherein h is c Namely the elevation of the feature point, is a fitting value generated by the common contribution of all three-dimensional point clouds in the search radius.
The invention discloses a preferable scheme of a target-free evaluation method for low-altitude unmanned point cloud modeling precision, which comprises the following steps: the chi-square fitness test, comprising,
eliminating abnormal coarse difference points according to the residual error of the point cloud and the reference surface, and performing chi-square fitting test between the three-dimensional point cloud and a fitting curved surface according to a data covariance matrix and a residual error vector r:
wherein the content of the first and second substances,
values of fitting parameters representing an adaptive polynomial fitting surface which gives a given χ of data points in a random Gaussian distribution 2 Values and corresponding hypothesis testing (P-value testing) probabilities.
The invention has the beneficial effects that: according to the invention, the target does not need to be laid during the operation of the unmanned aerial vehicle, the subsequent point cloud modeling precision evaluation can be completed, the time and labor for manually carrying the target in the field are avoided, and great convenience is brought to the operation of the unmanned aerial vehicle in the field.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise. Wherein:
fig. 1 is an overall flowchart of a low-altitude unmanned-machine point cloud modeling accuracy target-free evaluation method.
Fig. 2 is a flow chart of acquiring an unmanned aerial vehicle oblique photogrammetry image of the unmanned aerial vehicle point cloud modeling accuracy non-target evaluation method.
Fig. 3 is an interface diagram for setting parameters in the unmanned aerial vehicle flight operation of the low-altitude unmanned aerial vehicle point cloud modeling precision non-target evaluation method.
FIG. 4 is a two-dimensional model diagram of a test area of a low-altitude unmanned point cloud modeling accuracy target-free evaluation method.
Fig. 5 is a feature point diagram determined by a digital ortho-model DOM image based on the low-altitude unmanned point cloud modeling accuracy non-target evaluation method.
FIG. 6 is a three-dimensional model diagram of a test area of a low-altitude unmanned point cloud modeling accuracy target-free evaluation method.
Fig. 7 is a schematic diagram of a processing flow of an adaptive polynomial fitting surface of a low-altitude unmanned-machine point cloud modeling accuracy target-free evaluation method.
FIG. 8 is a self-adaptive polynomial fitting surface diagram of a low-altitude unmanned point cloud modeling accuracy target-free evaluation method.
Fig. 9 is a schematic diagram of elimination standards of rough difference points in a surface fitting process of the low-altitude unmanned point cloud modeling accuracy target-free evaluation method.
Fig. 10 is a chi-square detection schematic diagram of the low-altitude unmanned point cloud modeling accuracy target-free evaluation method.
Fig. 11 is a diagram of a test area of an unmanned aerial vehicle inclined three-dimensional laser scanning operation with a target.
Fig. 12 is a schematic diagram of target point positions of an unmanned aerial vehicle tilting three-dimensional laser scanning operation test with a target.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described and will be readily apparent to those of ordinary skill in the art without departing from the spirit of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.
Furthermore, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
Example 1
Referring to fig. 1 to 10, a first embodiment of the present invention provides a method for estimating the modeling accuracy of a low-altitude unmanned aerial vehicle point cloud without a target, including:
s1: collecting laser point cloud data of an unmanned aerial vehicle and measuring a control point;
it should be noted that the laser point cloud data required by the invention is collected by an unmanned aerial vehicle such as a Xinsi L1 radar lens carried by a Xinsi Longitude M300RTK unmanned aerial vehicle in Xinjiang. The specific flow of the laser point cloud oblique photogrammetry flight operation of the unmanned aerial vehicle is shown in fig. 2.
In step S1, the method further includes the steps of preparing before flight, setting an area on google earth as a general case of flight area, saving the KML file, and directly importing the KML file into the available flight designation area. And secondly, planning a route, and determining a survey area range after fully considering the surrounding environment factors through primary design and field survey planning. Then, according to the requirements of different target tasks, specific relevant parameters (refer to fig. 3) are set, including information of RTK and lens parameters, flight parameters, such as lens photography parameters, flight altitude, overlap degree, and the like, and in order to realize three-dimensional reconstruction, it is necessary to ensure that two adjacent flight maps overlap each other, and the overlap degree is usually set to be 60% -80%. On one hand, the larger the overlapping degree is, the denser the air route is, and the longer the corresponding flight time is; on the other hand, the more pictures taken, the more post-processing is performed. After the parameters are set, whether the unmanned aerial vehicle is connected with an RTK (real-time kinematic) network or not is confirmed, and the unmanned aerial vehicle can take off after the zen L1 lens completes inertial navigation preheating. And finally, processing the acquired data.
Furthermore, when the unmanned aerial vehicle operates, the control points are collected and measured, and reference is provided for subsequent point cloud modeling evaluation. The control points are selected strictly according to the following criteria: (1) when using an existing control point, the stability and integrity of the point should be checked; (2) the control point on the ground is selected to be in a place which is favorable for storage and convenient for measurement; (3) the control points should have a wide field of view and avoid the effects of multipath effects. After the control points are selected, a GNSS receiver such as a Wohmo i70 geodetic GNSS receiver is used for measuring and recording GPS coordinates and elevation information of the control points.
S2: acquiring a characteristic point GPS coordinate through two-dimensional modeling of unmanned aerial vehicle laser point cloud data;
the obtained two-dimensional modeling is shown in fig. 4, and in the step, high-precision DOM data can be obtained by using unmanned aerial vehicle mapping software such as a great-jiang wisdom diagram. In ArcGIS software, coordinates of DOM and control points are unified under a WGS 84/UTM Zone 50N coordinate system. Feature point positioning is then done based on the DOM data, see figure 5. Firstly, based on visual judgment, positioning macroscopic feature points, such as a well cover center, a flower bed corner point, a basketball court corner point, a zebra crossing corner point and the like, near a control point. And then, respectively making extension lines on two crossed boundaries of the characteristic ground object, so that the extension lines are crossed, and the intersection point is the characteristic point.
And calculating the horizontal distance deviation delta x and the vertical distance deviation delta y of the characteristic points and the control point GPS coordinates through pixel level comparison, and finally representing the modeled positioning error by using the straight line distance deviation delta l.
S3: acquiring a feature point elevation through three-dimensional modeling of unmanned aerial vehicle laser point cloud data;
the method comprises the following steps: converting the format of the three-dimensional model; intercepting point clouds in different ranges by taking the characteristic points as circle centers; using chi-square fitness test to remove point clouds with chi-square values larger than a critical value, and reducing gross errors; and performing self-adaptive polynomial fitting on the remaining point cloud to obtain elevation information of the characteristic points, and comparing the elevation information with the elevation information of the control points in actual measurement to represent modeling elevation errors.
Further, in the step, unmanned aerial vehicle mapping software such as a Dajiang intelligent map is used for carrying out laser point cloud three-dimensional modeling, and a high-precision three-dimensional model is generated, and reference is made to fig. 6. And (4) leading the model into point cloud processing software such as CloudCompare, and cutting and converting the model. And converting the three-dimensional model into a txt file attached with GPS information and elevation information, and spreading points in ArcGIS.
And (3) searching point clouds near the characteristic points by taking the characteristic points as circle centers and respectively taking different lengths such as 30cm, 40cm and 50cm as search radiuses l, and reconstructing the three-dimensional terrain surface by using a self-adaptive polynomial:
Z＝h c +a 1 x+a 2 y+a 3 x 2 +a 4 y 2 +a 5 xy+a 6 x 3 +a 7 y 3 +a 8 x 2 y+a 9 xy 2 (1)
considering that if the order of the polynomial is too high, a multi-fitting phenomenon of the curved surface is caused, and the error of the elevation estimation value is large, the order of the polynomial is determined to be 3, and 10 parameters are used for fitting the curved surface. Wherein x and y are coordinates of the point cloud under WGS 84/UTM Zone 50N coordinate system respectively. This polynomial can be converted to the following equation to solve for the 10 parameters:
wherein Z is the elevation of each point cloud, and e is the observation error, which is a range within
The value of (c). Sigma 0 Is unit weight, P is weight matrix of all point clouds in the search radius, according to the elevation and the feature point height of each point cloudThe error value of the process is obtained by giving different weights to each point cloud.
Further, B in formula (2) is a combination of polynomials of x and y in different orders, and can be represented by the following formula:
when the curve is fitted, the invention takes the characteristic point as an origin to establish a local coordinate system, and the coordinate of the characteristic point is expressed as (X) c ,Y c ) The coordinates of the other point clouds relative to the coordinates of the feature points are expressed as (X-X) c ,y-Y c ) And l is a search radius, and the division by a scaling coefficient (taking a search radius value as a reference) improves the least square solution precision and efficiency. m and n represent orders, and the highest order does not exceed 3. The values of m and n are determined according to the position distribution condition of the three-dimensional point cloud, if the point cloud is relatively wider than the characteristic points on the x coordinate, the order of x is increased, a better fitting result can be obtained, and the order of y is adjusted in the same way.
In the formula (2), X is a parameter equation to be solved, and can be represented by the following formula:
wherein h is c I.e., the elevation of the feature point, which is a fitting value generated by the common contribution of all three-dimensional point clouds in the search radius.
Because the point cloud contains noise points, all point clouds in the search radius can not be directly used for fitting a curved surface, and the gross error is removed by an effective method at this time. Therefore, the abnormal coarse difference point is eliminated according to the residual error of the point cloud and the reference surface. Performing chi-square fitting test between the three-dimensional point cloud and the fitting curved surface according to the data covariance matrix and the residual vector r:
the equation is used to express the consistency between the polynomial surface and the point cloud elevation, where
The fitting parameter value of the adaptive polynomial fitting surface is shown, which gives a given chi of a data point in random Gaussian distribution 2 Values and corresponding hypothesis testing (P-value testing) probabilities. If the probability is less than 0.025, half of the interval between the fitting residual probability distribution 16 and the 84 percentile value is a robust three-dimensional point RDE (robust Difference estimator), the RDE value which is 3 times of the robust three-dimensional point RDE is used as a rough Difference point rejection threshold, and the iteration termination minimum value is set to be 0.01m, so that the high-quality point is prevented from being rejected by mistake. The above process is repeated until terminated by hypothesis testing or iteration. And obtaining the high-precision central characteristic point elevation of the local surface of the three-dimensional point cloud. The adaptive polynomial fitting surface processing and the elimination process of rough difference points in the surface fitting process refer to fig. 7 to 10.
S4: and comparing the GPS coordinates of the characteristic points with the GPS coordinates of the control points, and comparing the elevation of the characteristic points with the elevation of the control points to finish the target-free evaluation.
It should be noted that step S1 is performed simultaneously with step S2, and not sequentially.
Example 2
In the second embodiment of the present invention, in order to better verify and explain the technical effects adopted in the method of the present invention, a conventional detection method with a labeled target is selected in the second embodiment, and the test results are compared by means of scientific demonstration to verify the real effects of the method;
the invention relates to a target-free evaluation method for low-altitude unmanned-machine point cloud modeling precision. In the feature point positioning of DOM data, the positioning accuracy of the low-altitude unmanned-machine point cloud modeling can be obtained by comparing the horizontal deviation and the vertical deviation of 30 control points and the feature points, calculating the straight line deviation and finally calculating RMSE.
Taking east lake school district of agriculture and forestry university of Zhejiang as an example, the unmanned aerial vehicle laser point cloud two-dimensional modeling is carried out, and the modeling positioning error is 0.168m through measurement and statistical analysis, and the test results are shown in the following table 1,
table 1: positioning error statistical table
Aiming at the three-dimensional point cloud model, taking the determined plane characteristic points as circle centers, respectively making circles with different radiuses such as 30cm, 40cm and 50cm, intercepting layer blocks in the three-dimensional model, outputting txt files of coordinates and elevation data of the layer blocks, and importing the layer blocks into an ArcGIS software exhibition point; based on point clouds generated in different ranges, point clouds with chi-square values larger than critical values are detected and removed by using chi-square fitness, then points with gross errors removed are used for conducting adaptive polynomial fitting to obtain an elevation curved surface, the optimal elevation of the characteristic points is obtained, and the GPS elevation of the control points and the elevation error of the characteristic points are evaluated. Similarly, taking east lake school district of agriculture and forestry university of Zhejiang as an example, performing unmanned aerial vehicle laser point cloud three-dimensional modeling, and obtaining that the elevation error of the modeling is 0.041m through measurement and statistical analysis, and the test results are shown in the following table 2,
table 2: elevation error statistical table
The elevation values of 30 characteristic points are obtained through adaptive polynomial surface fitting, more point clouds can be searched by using a search radius of 50cm through comparison, and the fitted elevation is closer to the elevation of the control point, so that the elevation precision of 0.04m is obtained by using the elevation fitted by using the search radius of 40 cm.
In comparison with the tilting three-dimensional laser scanning experiment of the unmanned aerial vehicle with the target, referring to fig. 11 and 12, the experimental results are shown in the following table 3,
table 3: unmanned aerial vehicle point cloud modeling error table with target
The target can be seen to be arranged in a large test area, the position of the target is required to be uniformly distributed, the sight line is required to be good, and the arrangement process consumes manpower and material resources. The non-target evaluation method for the low-altitude unmanned aerial vehicle point cloud modeling precision does not need to use a target, improves the convenience of unmanned aerial vehicle operation, and can obtain the same precision as that of unmanned aerial vehicle laser scanning operation with the target.
It is important to note that the construction and arrangement of the present application as shown in the various exemplary embodiments is illustrative only. Although only a few embodiments have been described in detail in this disclosure, those skilled in the art who review this disclosure will readily appreciate that many modifications are possible (e.g., variations in sizes, dimensions, structures, shapes and proportions of the various elements, values of parameters (e.g., temperatures, pressures, etc.), mounting arrangements, use of materials, colors, orientations, etc.) without materially departing from the novel teachings and advantages of the subject matter recited in this application. For example, elements shown as integrally formed may be constructed of multiple parts or elements, the position of elements may be reversed or otherwise varied, and the nature or number of discrete elements or positions may be altered or varied. Accordingly, all such modifications are intended to be included within the scope of this invention. The order or sequence of any process or method steps may be varied or re-sequenced according to alternative embodiments. In the claims, any means-plus-function clause is intended to cover the structures described herein as performing the recited function and not only structural equivalents but also equivalent structures. Other substitutions, modifications, changes and omissions may be made in the design, operating conditions and arrangement of the exemplary embodiments without departing from the scope of the present inventions. Therefore, the present invention is not limited to a particular embodiment, but extends to various modifications that nevertheless fall within the scope of the appended claims.
Moreover, in an effort to provide a concise description of the exemplary embodiments, all features of an actual implementation may not have been described (i.e., those unrelated to the presently contemplated best mode of carrying out the invention, or those unrelated to enabling the invention).
It should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort might be complex and time consuming, but would nevertheless be a routine undertaking of design, fabrication, and manufacture for those of ordinary skill having the benefit of this disclosure, without undue experimentation.
It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.
Claims (10)
1. A target-free evaluation method for low-altitude unmanned point cloud modeling precision is characterized by comprising the following steps: comprises the steps of (a) preparing a mixture of a plurality of raw materials,
collecting laser point cloud data of an unmanned aerial vehicle and measuring a control point;
acquiring a characteristic point GPS coordinate through two-dimensional modeling of unmanned aerial vehicle laser point cloud data;
acquiring a feature point elevation through three-dimensional modeling of unmanned aerial vehicle laser point cloud data;
and comparing the GPS coordinates of the characteristic points with the GPS coordinates of the control points, and comparing the elevation of the characteristic points with the elevation of the control points to finish the non-target evaluation.
2. The method for the non-target assessment of the modeling accuracy of a low altitude unmanned point cloud of claim 1, wherein: the collection of unmanned aerial vehicle laser point cloud data adopts unmanned aerial vehicle to carry on radar camera lens and gathers, and concrete step is:
preparing before flying, setting an area on virtual earth software, storing a KML file, and importing the KML file to obtain a flying designated area;
planning a route, namely determining a survey area range after considering surrounding environment factors through preliminary design and on-site survey planning;
setting specific parameters according to different target task requirements, wherein the parameters comprise RTK and lens parameters and flight parameters;
after the parameters are set, whether the unmanned aerial vehicle is connected with a network RTK or not is confirmed, whether the radar lens completes inertial navigation preheating or not is confirmed, and data collection is carried out after takeoff is confirmed.
3. The method for the drone-free assessment of low altitude drone point cloud modeling accuracy of claim 2, characterized by: the measurements of the control points, including,
selecting a control point, checking the stability and integrity of the existing control point, selecting a ground control point at a place where the control point is convenient to store and measure, and widening the visual field above the control point;
and measuring and recording the GPS coordinates and elevation information of the control points by using the GNSS receiver.
4. The method for the drone-free assessment of low altitude drone point cloud modeling accuracy of claim 3, characterized by: the two-dimensional modeling of the laser point cloud data of the unmanned aerial vehicle comprises the following specific steps:
adopting unmanned aerial vehicle mapping software to obtain high-precision DOM data;
unifying DOM data and control point coordinates to a WGS 84/UTM Zone 50N coordinate system in a GIS platform;
positioning a characteristic point based on DOM data, positioning a characteristic feature visible to the naked eye near a control point based on visual judgment in a geographic information system, and taking two crossed sidelines of the characteristic feature as an extension line, wherein the intersection point of the two extension lines is the characteristic point which is the position of the control point on a DOM image;
and calculating the horizontal distance deviation delta x and the vertical distance deviation delta y of the characteristic points and the control point GPS coordinates through pixel level comparison, and finally representing the modeled positioning error by using the straight line distance deviation delta l.
5. The method for the drone-free assessment of low altitude drone point cloud modeling accuracy of claim 4, characterized by: the three-dimensional modeling of the laser point cloud data of the unmanned aerial vehicle comprises the following specific steps:
creating a three-dimensional model through unmanned aerial vehicle mapping software;
converting the format of the three-dimensional model;
intercepting point clouds in different ranges by taking the characteristic points as circle centers;
using chi-square fitness test to remove point clouds with chi-square values larger than a critical value, and reducing gross errors;
and performing self-adaptive polynomial fitting on the remaining point clouds to obtain the elevation information of the characteristic points, comparing the elevation information with the elevation information of the control points in actual measurement, and representing the modeling elevation error by using delta h.
6. The method for the drone-free assessment of low altitude drone point cloud modeling accuracy of claim 5, characterized by: the format conversion of the three-dimensional model is specifically as follows:
and importing the generated three-dimensional model into point cloud processing software, performing cutting and data conversion on the three-dimensional model, converting the three-dimensional model into a txt file with GPS coordinates and elevation information, and importing the txt file into a GIS platform for point spread.
7. The method for the non-target assessment of the modeling accuracy of a low altitude unmanned point cloud of claim 6, wherein: intercepting point clouds in different ranges by taking the characteristic points as circle centers, specifically comprising the following steps:
the method comprises the following steps of searching point clouds near characteristic points by taking the characteristic points as circle centers and different lengths as search radiuses l, and reconstructing a three-dimensional terrain ground by using a self-adaptive polynomial, wherein the self-adaptive polynomial is as follows:
Z＝h c +a 1 x+a 2 y+a 3 x 2 +a 4 y 2 +a 5 xy+a 6 x 3 +a 7 y 3 +a 8 x 2 y+a 9 xy 2
wherein x and y are coordinates of the point cloud under a WGS 84/UTM Zone 50N coordinate system respectively, the order of the adaptive polynomial is determined to be 3, and 10 parameters are used for fitting the curved surface in order to avoid the multi-fitting phenomenon of the curved surface.
8. The method for drone-free assessment of low altitude drone point cloud modeling accuracy of claim 7, characterized by: the adaptive polynomial may be transformed to solve for 10 parameters as follows:
wherein Z is the elevation of each point cloud, and e is the observation error, which is a range within
The value of (c). Sigma 0 The unit weight is P, the weight matrix of all point clouds in the search radius is obtained by giving different weights to each point cloud according to the error value of the elevation of each point cloud and the elevation of the characteristic point;
b is a polynomial combination form of x and y with different orders, and is represented by the following formula:
when the curved surface is fitted, a local coordinate system is established by taking the characteristic points as the origin, and the coordinates of the characteristic points are expressed as (X) c ,Y c ) The coordinates of the other point clouds relative to the coordinates of the feature points are expressed as (X-X) c ,y-Y c ) L is a search radius, and the precision and the efficiency of least square solution are improved by dividing the search radius by a proportionality coefficient (taking a search radius value as reference);
m and n represent orders, and the highest order is not more than 3.
9. The method for drone-free assessment of low altitude drone point cloud modeling accuracy of claim 8, characterized by: in the conversion formula, X is a parametric equation to be solved, and is represented by the following formula:
wherein h is c Namely the elevation of the feature point, is a fitting value generated by the common contribution of all three-dimensional point clouds in the search radius.
10. The method for the non-target assessment of the modeling accuracy of a low altitude unmanned point cloud of claim 9, wherein: the chi-square fitness test, comprising,
eliminating abnormal coarse difference points according to the residual error of the point cloud and the reference surface, and performing chi-square fitting test between the three-dimensional point cloud and a fitting curved surface according to a data covariance matrix and a residual error vector r:
wherein the content of the first and second substances,
the fitting parameter values representing the adaptive polynomial fitting surface give the data points given χ in the random Gaussian distribution 2 Values and corresponding hypothesis testing (P-value testing) probabilities.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210387155.XA CN115018973A (en) | 2022-04-13 | 2022-04-13 | Low-altitude unmanned-machine point cloud modeling precision target-free evaluation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210387155.XA CN115018973A (en) | 2022-04-13 | 2022-04-13 | Low-altitude unmanned-machine point cloud modeling precision target-free evaluation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115018973A true CN115018973A (en) | 2022-09-06 |
Family
ID=83066389
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210387155.XA Pending CN115018973A (en) | 2022-04-13 | 2022-04-13 | Low-altitude unmanned-machine point cloud modeling precision target-free evaluation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115018973A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117456121A (en) * | 2023-10-30 | 2024-01-26 | 中佳勘察设计有限公司 | Topographic map acquisition and drawing method and device without camera |
-
2022
- 2022-04-13 CN CN202210387155.XA patent/CN115018973A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117456121A (en) * | 2023-10-30 | 2024-01-26 | 中佳勘察设计有限公司 | Topographic map acquisition and drawing method and device without camera |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102506824B (en) | Method for generating digital orthophoto map (DOM) by urban low altitude unmanned aerial vehicle | |
| CN108181635B (en) | Laser point cloud classification method for cross crossing analysis of power transmission line | |
| CN107063197B (en) | Reservoir characteristic curve extraction method based on spatial information technology | |
| CN104931022B (en) | Satellite image stereoblock adjustment method based on spaceborne laser altimeter system data | |
| El-Ashmawy | Investigation of the accuracy of google earth elevation data | |
| CN111724477A (en) | Method for constructing multi-level three-dimensional terrain model through multi-source data fusion | |
| CN106597416A (en) | Ground-GPS-assisted method for correcting error of difference of elevation of LiDAR data | |
| KR100510835B1 (en) | Method for constituting geographic information system applied digital map using real time measuring systems | |
| CN111006645A (en) | Unmanned aerial vehicle surveying and mapping method based on motion and structure reconstruction | |
| CN115018973A (en) | Low-altitude unmanned-machine point cloud modeling precision target-free evaluation method | |
| Gennaro et al. | Geomatic techniques for surveying and mapping an archaeological site | |
| CN115962755A (en) | Earth and stone calculation method based on unmanned aerial vehicle oblique photography technology | |
| Luh et al. | High resolution survey for topographic surveying | |
| Sefercik et al. | Point-based and model-based geolocation analysis of airborne laser scanning data | |
| Aguilar et al. | Assessing geometric reliability of corrected images from very high resolution satellites | |
| Taylor et al. | Surface modelling for GPS satellite visibility | |
| Kulur et al. | The Effect of Pixel Size on the Accuracy of Orthophoto Production | |
| Ping | Digital elevation model extraction from ASTER in support of the coal fire and environmental research project, China | |
| Hnila et al. | Quality Assessment of Digital Elevation Models in a Treeless High-Mountainous Landscape: A Case Study from Mount Aragats, Armenia | |
| Amran et al. | Horizontal coordinate accuracy of google earth on the coverage of small islands of Makassar City, Indonesia | |
| Varga et al. | Studying the accuracy of orthophotos on the example of various terrain models in study areas in Hungary | |
| Joshi et al. | Preparation of High-Resolution DTM and Orthophoto Using LiDAR in Nepal | |
| US20200111252A1 (en) | Computer storage means | |
| Sreedhar et al. | Line of sight analysis for urban mobile applications: a photogrammetric approach. | |
| Hnila et al. | A Case Study from Mount Aragats, Armenia |
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 |