CN111737882B - Landing zone selection method for realizing autonomous obstacle avoidance by complex lunar surface approaching section - Google Patents
Landing zone selection method for realizing autonomous obstacle avoidance by complex lunar surface approaching section Download PDFInfo
- Publication number
- CN111737882B CN111737882B CN202010707586.0A CN202010707586A CN111737882B CN 111737882 B CN111737882 B CN 111737882B CN 202010707586 A CN202010707586 A CN 202010707586A CN 111737882 B CN111737882 B CN 111737882B
- Authority
- CN
- China
- Prior art keywords
- landing zone
- annular mountain
- landing
- mountain
- annular
- 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.)
- Active
Links
- 238000010187 selection method Methods 0.000 title claims description 6
- 238000004458 analytical method Methods 0.000 claims abstract description 24
- 238000000034 method Methods 0.000 claims abstract description 17
- 230000003287 optical effect Effects 0.000 claims abstract description 14
- 230000005653 Brownian motion process Effects 0.000 claims abstract description 8
- 238000005537 brownian motion Methods 0.000 claims abstract description 8
- 238000013459 approach Methods 0.000 claims abstract description 7
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 claims description 11
- 238000004088 simulation Methods 0.000 claims description 9
- 238000003064 k means clustering Methods 0.000 claims description 7
- 239000000523 sample Substances 0.000 description 8
- 238000001514 detection method Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 5
- 238000012876 topography Methods 0.000 description 5
- 238000000926 separation method Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 241001631030 Explorator Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The method for selecting the landing zone for realizing autonomous obstacle avoidance in the complex lunar surface approach section solves the problem that the information utilization rate is not high enough when the autonomous obstacle avoidance and landing zone selection are carried out in the existing lunar soft landing, and belongs to the technical field of lunar soft landing. The method comprises the following steps: s1, acquiring an optical image of the lunar surface measured in a proximity section, and determining a landing zone to be selected; s2, extracting shape information of the annular mountain in the landing zone to be selected, reconstructing the annular mountain in the landing zone to be selected by combining with Fractal Brownian Motion (FBM), calculating gradients of points in the annular mountain, and obtaining a region with the gradient smaller than the safety gradient as a falling zone according to the set safety gradient; s3, performing texture analysis on the landing zone to be selected, and evaluating roughness to obtain a flat land block of the landing zone to be selected; s4, selecting the optimal feasible landing area by combining the lowerable area and the flat land block. The method provided by the invention can accurately reflect the shape characteristics of the annular mountain and the lunar surface in the lunar surface image, can effectively select the area suitable for landing, and has good engineering adaptability.
Description
Technical Field
The invention relates to a landing zone selection method for realizing autonomous obstacle avoidance by a complex lunar surface approach section, and belongs to the technical field of lunar soft landing.
Background
From the 60 s of the 20 th century, a number of lunar probes have been launched in countries and regions such as the united states, soviet union, europe, japan, china, india, etc. For early lunar exploration tasks, due to lack of autonomous obstacle avoidance capability, a relatively wide and flat place is usually planned as a task landing zone before task implementation, such as Luna series detectors and explorator series detectors; or the obstacle avoidance function is realized by means of manual operation, such as an Apollo series detector. To increase landing success rate, lunar probes are required to have the ability to autonomously select landing sites.
Since the 21 st century, the united states proposed an unmanned smart landing autonomous soft landing obstacle avoidance technique based on lidar, but according to the existing data, the technique has not been successfully applied. In the soft landing tasks of the 'goddess Chang' aircraft No. three and No. four, an obstacle avoidance algorithm combining coarse and fine is adopted to carry out autonomous obstacle avoidance and landing zone selection. Mainly performing rough obstacle avoidance on an approaching section (2400 m-100 m), resetting a guidance target and planning a landing area on the basis of optical information obtained by passive detection; in the hovering, obstacle avoidance and slow descending section (100 m-2 m), the lunar surface high-precision topographic information is obtained by adopting sensors such as a laser radar, a laser altimeter and the like, and the optimal landing point is selected to realize 'fine obstacle avoidance'.
The difficulty of the autonomous obstacle avoidance and landing zone selection technique for soft lunar landing is that the lack of lunar elevation map (Digital Elevation Model, DEM) information prior to task execution, and the inability to complete larger obstacle avoidance maneuvers when the detector obtains more accurate lunar elevation information at the end of landing. At present, based on passive optical detection information obtained by the approach section, obstacle identification and safe landing zone selection are mainly realized by analyzing the light and shade gray scale characteristics of obstacles such as annular mountains, stones and the like, and the information utilization rate is not high enough.
Disclosure of Invention
Aiming at the problem that the information utilization rate is not high enough when the existing lunar soft landing autonomous obstacle avoidance and landing zone selection are carried out, the invention provides a landing zone selection method for realizing autonomous obstacle avoidance by a complex lunar approaching section, which is used for roughly reconstructing the approaching section to be selected and recovering three-dimensional information from a two-dimensional image on the basis of fully analyzing the shape characteristics of the lunar annular mountain.
The invention discloses a landing zone selection method for realizing autonomous obstacle avoidance in a complex lunar surface approaching section, which comprises the following steps:
s1, acquiring an optical image of the lunar surface measured in a proximity section, and determining a landing zone to be selected;
s2, extracting shape information of the annular mountain in the landing zone to be selected, reconstructing the annular mountain in the landing zone to be selected by combining with Fractal Brownian Motion (FBM), calculating gradients of points in the annular mountain, and obtaining a region with the gradient smaller than the safety gradient as a falling zone according to the set safety gradient;
s3, performing texture analysis on the landing zone to be selected, and evaluating roughness to obtain a flat land block of the landing zone to be selected;
s4, selecting the optimal feasible landing area by combining the lowerable area and the flat land block.
Preferably, the S2 includes:
s21, extracting an annular mountain of a landing zone to be selected for circle fitting, and calculating the inter-pit-lip diameter D of the fitted annular mountain;
s22, obtaining the depth H and the pit lip height H of the annular mountain according to the formula I and the formula II r ：
H＝0.196D 1.01 Equation one
H r ＝0.036D 1.014 Formula II
S23, according to the inter-pit lip diameter D, depth H and pit lip height H of the annular mountain r Acquiring equivalent radius D of outer edge of pit lip r ：
S24, approximating the annular mountain to be a spherical surface, and calculating the equivalent spherical radius R of the annular mountain:
s25, according to the inter-pit-lip diameter D, the depth H and the pit-lip height H of the annular mountain r Equivalent radius D of the outer edge of the pit lip r And determining the shape of the annular mountain by using the equivalent spherical radius R, and overlapping the annular mountain by using the fractal-based Brownian motion FBM to form an annular mountain simulation point;
s26, calculating the gradient psi of each point in the annular mountain according to the simulation points of the annular mountain;
s27, according to the set safety gradient psi * And obtaining an area with the gradient smaller than the safety gradient as a falling area.
Preferably, the safety gradient ψ * ＝12。
Preferably, in the step S3:
s31, dividing the lunar surface optical image of the landing zone to be selected into lunar image textures of N categories according to different gray scale characteristics, wherein the gray scale characteristics are used for representing characteristics of a flat land block;
s32, performing texture analysis on each type of moon image texture in the landing zone to be selected by applying a K-means clustering method to obtain a clustering result, and determining the flat land block according to the clustering result.
Preferably, the N is 6,6 categories are respectively a flat lunar surface, a rough lunar surface, a large annular mountain concave surface, a large annular mountain convex surface, a small annular mountain concave surface and a small annular mountain convex surface.
The invention has the beneficial effects that: according to the invention, aiming at the gray level image obtained by optical detection of the detector, firstly, modeling is carried out on the annular mountain shape characteristic and the distribution characteristic obtained by detection, and the gradient of each point in the landing zone to be selected is calculated; then evaluating the roughness of the landing zone to be selected by adopting an image texture analysis method based on K-means clustering; considering gradient and roughness comprehensively, selecting a feasible landing zone in the landing zone to be selected, and the invention has three advantages: (1) The utilization rate of passive optical detection information is improved, and the accuracy of autonomous navigation and obstacle recognition is improved; (2) The obstacle structure in the invisible area can be restored, so that the autonomy of the lunar probe for selecting the landing area is obviously improved; (3) Provides more accurate and reliable initial information for the fine obstacle avoidance system of the goddess Chang' series detector. Simulation experiments are carried out on the method, and the experiment shows that the method provided by the invention can accurately reflect the shape characteristics of the annular mountain and the lunar surface in the lunar surface image, can effectively select the area suitable for landing, and has good engineering adaptability.
Drawings
FIG. 1 is a schematic flow chart of the present invention;
FIG. 2 is a circular mountain extraction and circle fitting;
FIG. 3 is an equivalent cross-sectional view of a circular mountain;
FIG. 4 is a three-dimensional shape fit of a circular mountain;
FIG. 5 is a ring mountain simulated point cloud;
FIG. 6 is a sample of an "goddess Chang E" image number one;
FIG. 7 is an equivalent slope of points in a circular mountain;
FIG. 8 is a texture analysis result;
FIG. 9 is a profile value distribution;
FIG. 10 is a graph showing the results of different category number profile values;
FIG. 11 is a flat lunar surface texture after expansion treatment;
FIG. 12 is an image taken by an goddess Chang' detector;
FIG. 13 is a region of landing area to be selected (region A);
FIG. 14 is a fractal Brownian motion gradient estimation;
FIG. 15 is a feasible landing zone selection (zone A);
FIG. 16 is a landing zone (zone B) to be selected;
FIG. 17 shows the result of the feasible landing zone selection (zone C).
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that, without conflict, the embodiments of the present invention and features of the embodiments may be combined with each other.
The invention is further described below with reference to the drawings and specific examples, which are not intended to be limiting.
In the moon soft landing task, autonomous navigation and obstacle avoidance are performed simultaneously, and the method mainly comprises the key technologies of autonomous navigation, obstacle recognition and landing point selection, obstacle relative navigation, obstacle avoidance guidance and control and the like. In this embodiment, taking an "goddess Chang" series aircraft as an example, after the attitude adjustment section, the "goddess Chang" series detector starts from the height of the approach section 1500m, the optical camera is approximately vertically pointed to the lunar surface, and the landing area to be selected is observed, so as to obtain the topographic obstacle information and perform rough obstacle avoidance. After the 'goddess Chang' detector III reaches 100m of the hovering section, the landing zone detection is completed within 30s by adopting a laser radar, a laser speedometer and other sensors, the safe landing point is selected and is mechanically lowered to 30m above the landing point, and the fine obstacle avoidance work of the obstacle avoidance section is completed. The landing time sequence of the detector No. four of the goddess Change is basically the same as that of the detector No. three of the goddess Change, but the detector No. three of the goddess Change completes the landing task under the guidance of the ultra-high resolution image acquired by the detector No. two of the goddess Change, and the detector No. four of the goddess Change completes the landing task under the condition of no imaging reference, the landing environment is complex and changeable, the configured sensor has longer acting distance and the approach section approximately drops vertically. The landing obstacle avoidance timing is shown in table 1.
TABLE 1 landing obstacle avoidance timing
This embodiment is directed to a viable landing zone selection for a proximity segment, comprising:
step one, acquiring an optical image of the lunar surface measured in a near section, and determining a landing zone to be selected according to a flight mission plan; in the first step, preprocessing such as denoising, binarization and the like is needed for the lunar surface optical image, so that the extraction of the annular mountain and the lunar surface texture analysis are facilitated.
Extracting shape information of an annular mountain in a landing zone to be selected, reconstructing the annular mountain in the landing zone to be selected by combining with Fractal Brownian Motion (FBM), calculating gradients of points in the annular mountain, and obtaining a region with the gradient smaller than the safety gradient as a falling zone according to the set safety gradient;
thirdly, performing texture analysis on the landing zone to be selected, and evaluating the roughness of the landing zone to be selected to obtain a flat land block of the landing zone to be selected;
and step four, selecting the optimal feasible landing area by combining the lowerable area and the flat land block.
This embodiment also gives an explanation about FBM-based lunar surface reconstruction:
natural topography of lunar surfaceCombining two-dimensional FBM features, a gaussian random process, corresponds to a random variable x=x (t) subject to N (0, t) α ) The random distribution, α, is the dimension and represents the roughness. The lunar surface topography statistical self-similarity, variation difference and section power spectral density respectively satisfy the following formulas (1) to (3).
p(X(t)＜x)＝p(X(γt)＜γ H t) (1)
E[X(t+h)-X(t)] 2 ＝k|h| 2H (2)
G(ω)＝2πkω -H (3)
Wherein: gamma is the number of iterative layers, k is the scale factor, and H is the irregular dimension. Fractal dimension F D The relation with H is [17] ：
F D +H＝3.
When the FBM is used for simulating the lunar surface, first, H (H is more than or equal to 0 and less than or equal to 1) is calculated by a first-order absolute moment estimation method.
(B in 4) H (t) is an FBM function, and the logarithm of the formula (4) is taken to obtain
Equation (5) approximates a linear relationship.
For the lunar surface two-dimensional FBM function B H (u), let u= (u) x ,u y ) For a point location vector, the form is as follows:
wherein:ω (x, y) is two-dimensional gaussian white noise. Moon surface dimension is calculated to "goddess Chang' second stereoscopic image reconstruction moon topography DEM data。
First, a lunar surface two-dimensional FBM function B is calculated H (u) incremental mean in two directions:
where N is the selected N data window width and delta is the incremental calculation step length, taking 50m. Calculating the overall average value:
taking the logarithm of the above formula according to the formula (5), obtaining:
log(E(Δ))＝log(E Const )+(3-F D )log(Δ) (10)
it can be derived that:
calculation of fractal dimension F from 5 200X 200 image regions taken in the existing data D The results were as follows:
TABLE 2 lunar surface fractal dimension
| Numbering device | 1 | 2 | 3 | 4 | 5 |
| Irregular dimension | 0.762 | 0.284 | 0.466 | 0.559 | 0.961 |
| Fractal dimension | 2.238 | 2.716 | 2.534 | 2.441 | 2.039 |
The fractal dimension of the selected five areas is larger than 2, which indicates that the lunar surface fractal characteristics are good, and the reconstruction can be carried out by adopting FBM. And (3) overlapping the lunar surface annular mountain information on the basis of reconstructing the FBM to complete the terrain simulation.
In a preferred embodiment, the second step of the present embodiment includes:
step two, extracting an annular mountain of a landing zone to be selected for circle fitting, and calculating the diameter D between pit lips of the fitted annular mountain, as shown in fig. 2;
step two, obtaining the depth H and the pit lip height H of the annular mountain according to the formulas (12) and (13) r ：
H＝0.196D 1.01 (12)
H r ＝0.036D 1.014 (13)
Wherein the distribution quantity rule of the annular mountain on the lunar surface can be represented by the following statistical model:
lgN(D)＝b×lg(D)+lg(a) (14)
wherein: d represents the diameter of the annular mountain, N (D) is the number of annular mountain with the diameter larger than D in unit area, and a and b are model parameters. Statistical distribution parameters of different sized annular mountains on the lunar surface and on the lunar sea are given as shown in Table 3 (units: meters).
TABLE 3 fractal distribution parameters for annular mountain
The relation between the height and the diameter of the lunar ring-shaped mountain accords with the empirical formulas (12) and (13) according to the statistics that the number of meteorites on the lunar surface is more than 33000.
Step two, three, the cross-sectional view of the annular mountain is shown in FIG. 3, according to the inter-pit lip diameter D, depth H and pit lip height H of the annular mountain r Acquiring equivalent radius D of outer edge of pit lip r ：
Step two, approximating the annular mountain to be a spherical surface, and calculating the equivalent spherical radius R of the annular mountain according to formulas (12) and (13):
the result of the circular mountain shape fitting is shown in fig. 4;
step two, according to the diameter D, depth H and pit-lip height H of the annular mountain between the pit lips r Equivalent radius D of the outer edge of the pit lip r And determining the shape of the annular mountain by using the equivalent spherical radius R, and overlapping the annular mountain by using the fractal-based Brownian motion FBM to form an annular mountain simulation point; on the basis of fig. 4, the annular mountain is overlapped by using the FBM, and an annular mountain simulation point cloud chart is formed as shown in fig. 5.
Step six, calculating the gradient psi of each point in the annular mountain according to the simulation points of the annular mountain:
wherein: (x) 0 ,y 0 ) The method is characterized in that the method is used for detecting the circle center of the annular mountain, and (x, y) is the coordinates of each point in the annular mountain.
Seventhly, according to the set safety gradient psi * And obtaining an area with the gradient smaller than the safety gradient as a falling area.
In the embodiment, the safety gradient psi of the rough obstacle avoidance feasible landing zone is set * =12. The closer to the edge of the annular mountain, the larger the slope of the terrain, so the threshold value psi of the obstacle avoidance slope is selected * Thereafter, the corresponding viable landing zone and annular in-mountain viable landing zone radius d may be determined, as shown in fig. 6 (a) - (d).
d＝R sin(ψ * ) (18)
In FIG. 6 (a), D is the diameter between the lips of the annular pit, D r The equivalent radius of the outer edge of the annular pit lip; the annular mountain depth H and the annular mountain equivalent spherical radius R can be calculated by the formula (12) and the formula (16), as shown in (c) of FIG. 6; after fitting the annular mountain, the gradient of each point in the annular mountain can be calculated by the formula (17), as shown in (d) of fig. 6; by selecting a safety gradient psi * The touchable area in the current parcel may be obtained as shown in fig. 6 (b).
The touchable area only meets the gradient requirement of landing in the landing zone, and the roughness of the land is considered simultaneously for selecting a feasible landing zone, and in a preferred embodiment, the third step of this embodiment includes:
dividing the lunar surface optical image of the landing zone to be selected into lunar image textures of N categories according to different gray scale characteristics, wherein the gray scale characteristics are used for representing the characteristics of a flat land block;
and thirdly, performing texture analysis on each moon image texture in the landing zone to be selected by applying a K-means clustering method to obtain a clustering result, and determining the flat land block according to the clustering result.
The K-means clustering algorithm is an unsupervised classification algorithm and can be used for carrying out texture analysis on the lunar surface optical detection image. The texture analysis is used for evaluating the roughness of the landing zone to be selected, selecting a flatter land block, taking the area with smaller gradient and flatter gradient as a feasible landing zone, and improving the reliability of autonomous obstacle avoidance and soft landing.
In the preferred embodiment, N is 6,6 categories are flat lunar surface, rough lunar surface, large annular mountain concave surface, large annular mountain convex surface, small annular mountain concave surface, and small annular mountain convex surface, respectively.
Corresponding to 6 kinds of tagged data clusters Ci (i=1, 2, …, 6) as shown in table 4. The K-means clustering algorithm is applied to separate the images for texture analysis. Let the gray matrix of the picture be a label-free dataset, with n points.
X＝[x (1) x (2) ... x (n) ] T (19)
C＝[C 1 ,C 2 ,C 3 ,C 4 ,C 5 ,C 6 ] (20)
Table 4 moon surface image texture classification
The images are divided into 6 clusters according to the classification number k=6, and a minimization loss objective function J is constructed:
wherein: mu (mu) i Is cluster C i Is defined by a center point of (2); ||x i -μ i || 2 Is cluster C i The distance between each point in (a) and the center point.
And selecting a moon image sample shot by the goddess Chang E detector I for texture analysis, wherein the image sample is shown in figure 7.
K-means cluster texture analysis is performed on FIG. 7, and the analysis results are shown in FIG. 8.
In fig. 8, (a) to (f) are the same areas, and the corresponding white areas in the plots are selected and eligible plots. In fig. 8 (a) the white area is the smaller annular mountain cloudy surface; in fig. 8 (b), the white area is a larger annular mountain cloudy surface; in fig. 8 (c), the white area is a rough lunar surface; in fig. 8 (d), the white area is a flat moon surface; in fig. 8 (e) the white area is the smaller annular mountain surface; in fig. 8 (f), the white area is a larger annular mountain surface.
After the clustering result is obtained, the clustering result is evaluated by adopting a contour coefficient S (Silhouette Coefficient). For any point i in the gray scale image: the dissimilarity a (i) in the cluster represents the aggregation degree of the cluster, and a (i) is the average value of distances from i to other points in the same cluster; the inter-cluster dissimilarity b (i) represents the degree of separation of the cluster, and b (i) is the minimum value of the distances from i to the centers of other clusters. The i-point profile coefficient S (i) is:
the profile coefficient S of the cluster is:
the closer S is to 1, the better the clustering result is, the higher the aggregation degree of the data clusters is, and the more different data clusters are scattered; when S (i) <0, it is shown that the clustering of the points is not reasonable, and the points should be divided into other data clusters. The clustering results of fig. 7 and 8 were evaluated, and fig. 7 was classified into 4 to 7 categories, and the profile value distribution was calculated, and the result is shown in fig. 9.
When the number of the clustering categories is 6, the number of gray sample points with negative contour values is minimum, which indicates that the clustering data has good cluster separation degree and high cluster separation degree when being divided into 6 categories.
The corresponding profile values for the different category numbers are shown in fig. 10.
Considering the completeness of the annular mountain in the lunar surface texture analysis, the clustering type is excessive when the clustering type is 8, and the region where the same annular mountain is located is easily divided into a plurality of two parts, so that the complete annular mountain is not easy to identify; when the separation number is too small, the target topography meeting the landing condition cannot be extracted, so that the clustering number is selectable to be 4-7, and the corresponding clustering contour value and clustering duration are shown in table 5:
table 5 clustering profile values and clustering duration
When the clustering number is 6, the clustering contour value is the largest, which indicates that the proposed moon image texture analysis based on K-means clustering can classify moon gray images according to task requirements, distinguish different image textures to correspond to actual topography, and provide support for feasible landing zone selection.
According to the analysis result of the lunar image texture, selecting the flat lunar surface corresponding to the texture (d) in fig. 8 as a feasible landing area. In fig. 11 (a) is an original image of a flat lunar surface in fig. 8 (d), the white area is expanded to form the white area in fig. 11 (b), and the roughness of the white area in fig. 11 (b) is smaller, so that it can be seen that all annular mountains are basically avoided and landing is possible.
And (3) experimental verification:
the lunar image photographed by the goddess Chang' first detector is selected as an analysis sample, and as shown in fig. 12 (a), the available landing zone selection is performed based on fractal characteristics and lunar texture analysis, respectively.
First, gaussian blur and binarization are performed on the original image, and the edge and diameter of the annular mountain in the image are recognized, as shown in fig. 12 (b).
According to the task planning requirement, a certain area (area A) is taken as a landing area to be selected in the area (a) in fig. 12, as shown in fig. 13.
Reconstructing the terrain of fig. 13 through the FBM and calculating the slope of the terrain at each point; the safe landing zone slope was set to 12 deg., and plots with a slope greater than the safe slope were obtained as shown in the black area of fig. 14.
In fig. 14, the black area is an area with a larger average gradient, and cannot be used as a viable landing zone. Meanwhile, the texture analysis is carried out on the graph 13, the landing feasibility is evaluated by comprehensively considering the gradient and the roughness of the image area, and a feasible landing area is selected as shown in the graph 15:
in fig. 12, an alternative landing zone (zone B) is taken for analysis, as in fig. 16 (a); the landing feasibility was evaluated, and the feasible landing zone was obtained as shown in fig. 16 (b):
in fig. 16, a feasible landing zone is selected by considering the result of image texture analysis and the estimated gradient of FBM, and a "black" flat area is left in the upper left corner, but is not selected as the feasible landing zone, because the area is located inside the annular mountain and the gradient is also larger.
In fig. 12, an area (area C) with uneven gray value distribution of the image is alternatively selected, and landing feasibility is evaluated, so that a feasible landing area cannot be obtained, as shown in fig. 17.
It can be seen that the invention can reconstruct the terrain, identify the obstacle and autonomously select the effective feasible landing zone on the basis of reconstructing the lunar surface aiming at different landing zones to be selected.
Although the invention herein has been described with reference to particular embodiments, it is to be understood that these embodiments are merely illustrative of the principles and applications of the present invention. It is therefore to be understood that numerous modifications may be made to the illustrative embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the different dependent claims and the features described herein may be combined in ways other than as described in the original claims. It is also to be understood that features described in connection with separate embodiments may be used in other described embodiments.
Claims (2)
1. The method for selecting the landing zone for realizing autonomous obstacle avoidance by using the complex lunar surface approach section is characterized by comprising the following steps of:
s1, acquiring an optical image of the lunar surface measured in a proximity section, and determining a landing zone to be selected;
s2, extracting shape information of the annular mountain in the landing zone to be selected, reconstructing the annular mountain in the landing zone to be selected by combining with Fractal Brownian Motion (FBM), calculating gradients of points in the annular mountain, and obtaining a region with the gradient smaller than the safety gradient as a falling zone according to the set safety gradient;
s3, performing texture analysis on the landing zone to be selected, and evaluating roughness to obtain a flat land block of the landing zone to be selected;
s4, selecting an optimal feasible landing zone by combining the lowerable zone and the flat land block;
the step S2 comprises the following steps:
s21, extracting an annular mountain of a landing zone to be selected for circle fitting, and calculating the inter-pit-lip diameter D of the fitted annular mountain;
s22, obtaining the depth H and the pit lip height H of the annular mountain according to the formula I and the formula II r ：
H＝0.196D 1.01 Equation one
H r ＝0.036D 1.014 Formula II
S23, according to the inter-pit lip diameter D, depth H and pit lip height H of the annular mountain r Acquiring equivalent radius D of outer edge of pit lip r ：
S24, approximating the annular mountain to be a spherical surface, and calculating the equivalent spherical radius R of the annular mountain:
s25, according to the inter-pit-lip diameter D, the depth H and the pit-lip height H of the annular mountain r Equivalent radius D of the outer edge of the pit lip r And determining the shape of the annular mountain by using the equivalent spherical radius R, and overlapping the annular mountain by using the fractal-based Brownian motion FBM to form an annular mountain simulation point;
s26, calculating the gradient psi of each point in the annular mountain according to the simulation points of the annular mountain;
s27, according to the set safety gradient psi * Obtaining a region with gradient smaller than the safety gradient as a falling region;
in the step S3:
s31, dividing the lunar surface optical image of the landing zone to be selected into lunar image textures of N categories according to different gray scale characteristics, wherein the gray scale characteristics are used for representing characteristics of a flat land block;
s32, performing texture analysis on each type of moon image texture in the landing zone to be selected by applying a K-means clustering method to obtain a clustering result, and determining a flat land block according to the clustering result;
the N is 6,6 categories are respectively a flat lunar surface, a rough lunar surface, a large annular mountain concave surface, a large annular mountain convex surface, a small annular mountain concave surface and a small annular mountain convex surface.
2. The landing zone selection method for achieving autonomous obstacle avoidance in a complex lunar surface approach segment of claim 1, wherein the safety grade ψ is * ＝12。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010707586.0A CN111737882B (en) | 2020-07-21 | 2020-07-21 | Landing zone selection method for realizing autonomous obstacle avoidance by complex lunar surface approaching section |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010707586.0A CN111737882B (en) | 2020-07-21 | 2020-07-21 | Landing zone selection method for realizing autonomous obstacle avoidance by complex lunar surface approaching section |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111737882A CN111737882A (en) | 2020-10-02 |
| CN111737882B true CN111737882B (en) | 2024-02-27 |
Family
ID=72656233
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010707586.0A Active CN111737882B (en) | 2020-07-21 | 2020-07-21 | Landing zone selection method for realizing autonomous obstacle avoidance by complex lunar surface approaching section |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111737882B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114359320B (en) * | 2021-12-15 | 2023-02-03 | 哈尔滨工业大学 | Moon detector robust ring mountain detection method and aircraft navigation method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102173313A (en) * | 2010-12-24 | 2011-09-07 | 北京控制工程研究所 | Soft landing relay obstacle avoiding method |
| CN106527473A (en) * | 2016-10-27 | 2017-03-22 | 上海航天控制技术研究所 | Obstacle-avoidance landing method on lunar surface |
| CN110647159A (en) * | 2019-09-23 | 2020-01-03 | 北京控制工程研究所 | Precise obstacle avoidance heterogeneous backup method for lunar soft landing |
-
2020
- 2020-07-21 CN CN202010707586.0A patent/CN111737882B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102173313A (en) * | 2010-12-24 | 2011-09-07 | 北京控制工程研究所 | Soft landing relay obstacle avoiding method |
| CN106527473A (en) * | 2016-10-27 | 2017-03-22 | 上海航天控制技术研究所 | Obstacle-avoidance landing method on lunar surface |
| CN110647159A (en) * | 2019-09-23 | 2020-01-03 | 北京控制工程研究所 | Precise obstacle avoidance heterogeneous backup method for lunar soft landing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111737882A (en) | 2020-10-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110264572B (en) | Terrain modeling method and system integrating geometric characteristics and mechanical characteristics | |
| Yu et al. | A new approach based on crater detection and matching for visual navigation in planetary landing | |
| CN109558859B (en) | Mining area distribution information extraction method and system based on DInSAR and DCNN | |
| CN103268358B (en) | Multi-source control point image database builds and update method | |
| Chen et al. | Lunar crater detection based on terrain analysis and mathematical morphology methods using digital elevation models | |
| CN103218787B (en) | Multi-source heterogeneous remote sensing image reference mark automatic acquiring method | |
| Westoby et al. | Interannual surface evolution of an Antarctic blue-ice moraine using multi-temporal DEMs | |
| CN105184250A (en) | Ground object classification method based on electric power corridor airborne LiDAR point cloud data | |
| Ayhan et al. | Semi-automated emergency landing site selection approach for UAVs | |
| Shen et al. | A new algorithm of building boundary extraction based on LIDAR data | |
| KR101549155B1 (en) | Method of automatic extraction of building boundary from lidar data | |
| Hui et al. | An active learning method for DEM extraction from airborne LiDAR point clouds | |
| CN115761303A (en) | Ground object classification method based on airborne laser radar point cloud and remote sensing image data | |
| CN111737882B (en) | Landing zone selection method for realizing autonomous obstacle avoidance by complex lunar surface approaching section | |
| CN107292039B (en) | UUV bank patrolling profile construction method based on wavelet clustering | |
| Wang et al. | Automatic mapping of lunar landforms using DEM-derived geomorphometric parameters | |
| Savage et al. | A Bayesian approach to subkilometer crater shape analysis using individual HiRISE images | |
| CN111721302B (en) | Method for recognizing and sensing complex terrain features on surface of irregular asteroid | |
| CN107742295A (en) | A kind of cube star docking reconstructing method of view-based access control model | |
| Yin et al. | A novel method of crater detection on digital elevation models | |
| Lee et al. | Determination of building model key points using multidirectional shaded relief images generated from airborne LiDAR data | |
| CN112906573B (en) | Planet surface navigation road sign matching method based on contour point set | |
| CN115511899A (en) | Method and device for extracting morphological structure of shallow small landslide | |
| Wu et al. | Robust crater detection algorithm based on maximum entropy threshold segmentation | |
| Sayed et al. | Point clouds reduction model based on 3D feature extraction |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |