CN114187529B - Small celestial body surface complex terrain feature detection method - Google Patents
Small celestial body surface complex terrain feature detection method Download PDFInfo
- Publication number
- CN114187529B CN114187529B CN202111472200.3A CN202111472200A CN114187529B CN 114187529 B CN114187529 B CN 114187529B CN 202111472200 A CN202111472200 A CN 202111472200A CN 114187529 B CN114187529 B CN 114187529B
- Authority
- CN
- China
- Prior art keywords
- region
- area
- edge
- shadow
- image
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 46
- 238000000034 method Methods 0.000 claims abstract description 24
- 238000012876 topography Methods 0.000 claims abstract description 13
- 238000003708 edge detection Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000005286 illumination Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 9
- 238000003384 imaging method Methods 0.000 claims description 7
- 230000003287 optical effect Effects 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 4
- 230000004069 differentiation Effects 0.000 claims description 3
- 230000008030 elimination Effects 0.000 claims description 3
- 238000003379 elimination reaction Methods 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 2
- 238000011160 research Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2133—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on naturality criteria, e.g. with non-negative factorisation or negative correlation
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Automation & Control Theory (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for detecting complex topography features of the surface of a celestial body, and belongs to the technical field of deep space exploration. The implementation method of the invention comprises the following steps: the navigation image shot by the detector is utilized, the shadow area and the bright area of the navigation feature of the surface of the celestial body are obtained by extracting the feature of the maximum stable extremum area and according to the gray level average value of the area, and the repeated area is removed by utilizing the distance constraint to merge the maximum stable extremum area; and setting the size of an initial pairing search window according to the number of pixels in the shadow area, and realizing the rough extraction of the edges of the same navigation feature. Aiming at the problem of false detection of the edge of the meteorite crater in the overlapped area, the number of the public edge points searched by the two shadow areas is used as feedback information, a geometric mean square distance minimum principle is established to realize dynamic adjustment of the searching radius, and accurate detection of the features of the meteorite crater in the overlapped area is realized. The invention can provide an accurate autonomous detection method for deep space exploration of complex terrain features, thereby providing accurate and reliable navigation roadmap for the navigation system of the detector.
Description
Technical Field
The invention relates to a method for detecting complex topographic features of the surface of a small celestial body, in particular to a method for extracting complex topographic features of overlapped meteorite pits and the like of the surface of the small celestial body, and belongs to the technical field of deep space detection.
Background
In the small celestial body detection task, because the small celestial body is far away from the earth, a large measurement delay exists by utilizing a ground station measurement and control mode, and the navigation instantaneity and reliability are difficult to guarantee. With the breakthrough of computer hardware technology and the development of optical sensitive devices, an optical autonomous navigation method based on planetary surface features (merle pits and rocks) becomes a research hotspot. The merle pit is used as a terrain feature of the surface of the small celestial body, has a relatively consistent geometric shape and a clear outline, is little influenced by illumination change, and is an ideal landing navigation terrain feature. Considering that the area with scientific research value on the surface of the celestial body is often complex in topography, how to accurately identify complex topography features is a key for realizing high-precision autonomous navigation of the detector.
In the proposed merle feature recognition detection method, the prior art (Meng Y,Cui H,Yang T.A new approach based on crater detection and matching for visual navigation in planetary landing[J].Advances in Space Research,2014,53(12):1810-1821.) proposes a regional-based merle detection algorithm, which uses the extracted shadow region as the center, determines the search radius by setting an empirical parameter, and finds the bright region belonging to the same merle, thereby realizing the merle detection. Because the method uses empirical parameters to set the search radius, the overlapping merle features cannot be accurately identified. Therefore, research is necessary to develop a method for detecting the complex terrain features of the surface of the celestial body so as to realize high-precision navigation of the detector under the complex terrain.
Disclosure of Invention
The invention discloses a method for detecting complex terrain features of the surface of a celestial body, which aims to solve the problems that: aiming at complex topography of the surface of the small celestial body, extracting the edge and the area of a navigation image shot by a detector, merging the maximum stable extremum area by utilizing distance constraint to remove the repeated area, aiming at the problem of false detection of the edge of the merle pit in the overlapped area, taking the number of public edge points searched by the shadow area as feedback information, establishing a minimum principle of geometric mean square distance to realize dynamic adjustment of the searching radius, and realizing accurate detection of the feature of the merle pit with the overlapped area.
The invention is realized by the following technical scheme.
According to the method for detecting the complex topography features of the surface of the celestial body, disclosed by the invention, the shadow area and the light area of the navigation features of the surface of the celestial body are obtained by extracting the maximum stable extremum region features and according to the region gray level average value by utilizing the navigation images shot by the detector. And setting the size of an initial pairing search window according to the number of pixels in the shadow area, and realizing the rough extraction of the edges of the same navigation feature. Aiming at the problem of false detection of the edge of the merle pit in the overlapped area, the number of the public edge points searched by the two shadow areas is used as feedback information, dynamic adjustment of the searching radius is realized, a fitting principle with minimum geometric mean square distance is established, ellipse fitting is carried out on the extracted result, a best fitting scheme is obtained, and an accurate autonomous detection method is provided for deep space exploration of complex terrain features.
The invention discloses a method for detecting complex topography characteristics of the surface of a celestial body, which comprises the following steps:
Step 1: and (3) performing image processing on the image shot by the navigation optical camera, extracting edge and region information, merging the maximum stable extremum regions by using distance constraint to remove repeated detection results of the same region, and extracting a shadow region of the navigation features of the surface of the celestial body according to the gray average value of the region.
The specific implementation method of the step 1 is as follows:
Step 1.1, edge detection is carried out on the merle pit on the surface of the celestial body in the navigation camera image, so that edge information of the image is obtained, and short edges and pseudo edges are removed.
Based on a canny edge detection algorithm, edge detection is carried out on the merle pit on the surface of the celestial body, short edges are removed through the communication area marks, the extracted binary image edge information is marked as B (u, v), the corresponding gradient vector is marked as g (u, v), and u and v respectively represent the horizontal coordinates and the vertical coordinates of edge pixel points in the image. The gradient direction of the false edge detected in the meteorite crater is opposite to that of the true edge of the meteorite crater, and the elimination of the false edge features can be realized through the illumination direction. The included angle between the gradient direction of the meteorite crater edge and the direction of the light source is smaller than 90 degrees, and for the image gradient vector g (u, v) of the edge and the direction vector n of the light source on the image plane, the true edge of the meteorite crater satisfies:
And (3) realizing false edge elimination by using constraint shown in a formula (1) to obtain edge information of the image.
And 1.2, extracting the maximum stable extremum region of the navigation image, utilizing the distance constraint to realize region merging and removing repeated detection results of the same topographic feature, and extracting the shadow region of the celestial body surface navigation feature according to the region gray average value.
Based on the MSER characteristic of the maximum stable extremum region, image processing is carried out on the merle pit on the surface of the celestial body, the region characteristic of the image is obtained, and the merle pit image is initially divided into a shadow region image and a bright region image. The basic principle of the MSER is that the gray level images are sequentially subjected to binarization processing by taking increasing thresholds, and in all obtained binary images, the area with small change of the connected area is called as the maximum stable extremum area. And (3) marking the center of the obtained MSER characteristic as L (u, v), carrying out region merging on the obtained region characteristic through the constraint condition shown in the formula (2) by virtue of the distance constraint so as to remove repeated detection results of the same characteristic, and extracting a shadow region of the celestial body surface navigation characteristic according to the region gray average value.
L(ui,vi)-L(uj,vj)＞ε (2)
Wherein u i、vi represents the abscissa of the center of the ith feature region in the image, ε is a user-defined parameter, and the features of the same region are considered without satisfying constraint (2), and the two are combined. The region is divided into a shadow region and an illumination region by the region gray average value, the binary image of the shadow region of the extracted navigation feature is marked as B S (u, v), and the number of the shadow regions is marked as n S.
Step 2: according to the shape characteristics of the shadow region, setting the size of an initial search window, carrying out edge search on the navigation features to obtain detected meteorite pit edges, carrying out ellipse fitting on the detected meteorite pit edges based on the geometric mean square distance minimum principle, realizing the preliminary detection of the navigation features, and obtaining the geometric mean square error GRMSE of each fitted ellipse.
And (3) because the shadow region extracted in the step (1) is elliptical, the bright region is non-elliptical, elliptical fitting is carried out on the shadow region according to the imaging characteristics of the region shape, and the center of the ellipse obtained by fitting is taken as the center of the shadow region and is marked as B C＝(uc,vc). Setting the size of a search window according to the size of the shadow area by taking the center of the shadow area as the center of the search window, and searching all edges of the same meteorite pit on the edge detection result in the step 1. The purpose of the local search is to reduce the calculation amount of global search imaging matching and improve the search efficiency.
The search window is a circular area with the center B C of the shadow area as the center and R as the radius, all edges of the same meteorite pit are searched in the area, the size of the circle is defined as k times of the total number of pixels of the shadow area, and the calculation formula of the search radius R n of the nth shadow area is as follows:
Where k is a custom search parameter. D A is the number of pixels contained in the nth shaded region.
And carrying out ellipse fitting based on a geometric mean square distance minimum principle on the detected meteorite pit edge, realizing the preliminary detection of navigation features, and obtaining the geometric mean square error GRMSE of each fitted ellipse, wherein the calculation steps are as follows.
The fitting approximate mean square error of the two-dimensional curve is:
q is the number of two-dimensional point concentration points, f is described by a finite parameter, denoted as f (x) ≡phi (α, x), α is a smooth function, for a particular α= (α 1···αr)T, f can be written as f=Φ α (x), α is referred to as a parameter, and x is referred to as a variable.
In ellipse fitting, the minimization problem of equation (4) can be reduced to a generalized eigenvector problem, where equation (4) becomes:
In the middle of Is the covariance matrix and trace represents the trace of the matrix.
The fit constraint is:
In the middle of Is a symmetric non-negative definite matrix, D represents the differentiation of the function.
Ellipse fitting is the computation of discrete pointsAnd/>And solving the generalized eigenvector problem m·c=λn·c, the minimum eigenvalue λ and the corresponding eigenvector c gives a solution to the elliptical condition. λ is the geometric mean square error and c is the coefficient of a quadratic equation representing the optimal ellipse.
Step 3: aiming at the problem of false detection of the edge of the merle pit in the overlapped area, the number of the public edge points searched by the shadow area is used as feedback information to realize dynamic adjustment of the searching radius, and ellipse fitting is carried out on the characteristic edge again until the obtained geometric mean square error is minimum, so that accurate detection of the feature of the merle pit in the overlapped area is realized.
The dynamic adjustment principle of the search radius refers to: and (3) adjusting the searching radius R n according to the number of the public edge points searched by the shadow region until the geometric mean square error GRMSE of each fitting ellipse obtained in the step (2) reaches the minimum value, wherein the specific implementation mode is as follows.
Through the ellipse preliminary detection of the step2, most merle pits on the surface of the celestial body can be successfully identified, but overlapping merle pits can search the edges of adjacent features when edge searching is carried out, and the fitting precision is poor. Recording the common point searched by two overlapped meteorites as P n, constructing an adjustment coefficientAnd dynamically adjusting R n until the optimal solution of the ellipse fitting is found. Beta is a scaling factor between 0 and 1, and the search radius/>The ellipse fitting problem thus far translates into a determination of α.
Written as constraint problem as shown in equations (7) (8) (9):
and (3) taking the ellipse corresponding to the minimum value as the optimal ellipse in the formula (7), and realizing the accurate detection of the merle pit characteristics with the overlapped area.
The beneficial effects are that:
1. Aiming at the problem of false detection of the edge of the merle pit in the overlapped area, the method for detecting the complex topography features of the surface of the small celestial body disclosed by the invention takes the number of the public edge points searched by the shadow area as feedback information, realizes dynamic adjustment of the searching radius, establishes the minimum geometric mean square distance principle to obtain the optimal result of ellipse fitting, realizes accurate detection of the feature of the merle pit in the overlapped area, and provides accurate and reliable navigation road signs for a navigation system of a detector.
2. According to the method for detecting the complex terrain features on the surface of the celestial body, disclosed by the invention, the maximum stable extremum region is combined by utilizing the distance constraint, and the repeated detection result of the same feature is effectively eliminated.
3. Due to the complex morphology of the surface of the celestial body, the overlapping condition of a plurality of merle pits can occur, and the method for detecting the complex topography features of the surface of the celestial body can realize the dynamic adjustment of each feature searching radius through the common points obtained by the detection of every two merle pits, so that the method is suitable for the overlapping condition of a plurality of merle pits, and the detection rate of navigation road signs of the surface of the celestial body is improved.
Drawings
FIG. 1 is a schematic flow chart of a method for detecting the navigation features of the surface of a celestial body;
FIG. 2 is an original navigation image taken by a deep space probe used for simulation in an example of the present invention;
FIG. 3 is a schematic image of the shadow and light fields in the merle in an example of the invention;
FIG. 4 is a diagram of edge detection by step 1 in an example of the present invention, wherein FIG. 4 (a) is a diagram after small area region removal, and FIG. 4 (b) is an edge detection result after false edge removal;
FIG. 5 is the meteorite crater shadow region extracted in step 1 of the present example;
FIG. 6 is a graph of the results of step 2 navigation feature edge ellipse fitting in an example of the present invention;
FIG. 7 is a graph of the results of an ellipse fit after step 3 dynamic adjustment of the search radius in an example of the present invention.
Detailed Description
For a better description of the objects and advantages of the present invention, the following description will be given with reference to the accompanying drawings and examples.
In order to verify the feasibility of the invention, mathematical simulation verification is performed by using a truly photographed planet surface merle image, as shown in fig. 2.
As shown in fig. 1, the method for detecting the complex topography characteristics of the surface of the celestial body disclosed in the embodiment specifically comprises the following implementation steps:
Step 1: and (3) performing image processing on the image shot by the navigation optical camera, extracting edge and region information, merging the maximum stable extremum regions by using distance constraint to remove repeated detection results of the same region, and extracting a shadow region of the navigation features of the surface of the celestial body according to the gray average value of the region.
The specific implementation method of the step 1 is as follows:
Step 1.1, edge detection is carried out on the merle pit on the surface of the celestial body in the navigation camera image, so that edge information of the image is obtained, and short edges and pseudo edges are removed.
The image of the surface topography of the target celestial body shot by the navigation optical camera is shown as a figure 2, and edge detection is carried out on the meteorite crater on the surface of the celestial body based on a canny edge detection algorithm. The 8 connected regions are obtained from the obtained edge detection result, short edges are removed, as shown in fig. 4 (a), the edge information of the extracted binary image is marked as B (u, v), the corresponding gradient vectors are marked as g (u, v), and u and v respectively represent the abscissa and ordinate of the edge pixel point in the image. Under illumination conditions, the imaging characteristics of the meteorite craters are shown as shown in figure 3, because the meteorite craters are bowl-shaped characteristics, a light bright area and a shadow area can appear in sequence in the illumination direction, the boundary between the light bright area and the shadow area is a pseudo edge characteristic, the gradient direction of the pseudo edge detected in the meteorite craters is opposite to the gradient direction of the true edge of the meteorite craters, and the pseudo edge characteristic can be removed through the illumination direction. The angle between the gradient direction of the merle edge and the direction of the light source is less than 90 degrees. The image gradient vector g (u, v) of the edge and the direction vector n of the light source on the image plane satisfy the following conditions:
False edge culling is achieved using the constraints described above, the result being shown in fig. 4 (b).
And 1.2, extracting the maximum stable extremum region of the navigation image, utilizing the distance constraint to realize region merging and removing repeated detection results of the same topographic feature, and extracting the shadow region of the celestial body surface navigation feature according to the region gray average value.
Based on the Maximum Stable Extremum Region (MSER) characteristic, image processing is carried out on the merle on the surface of the celestial body, the regional characteristic of the image is obtained, and the merle image is initially divided into a shadow region image and a bright region image. The basic principle of the MSER is that the gray level images are sequentially subjected to binarization processing by taking increasing thresholds, and in all obtained binary images, the area with small change of the connected area is called as the maximum stable extremum area. The center of the obtained MSER feature is marked as L (u, v), the obtained region features are subjected to region merging through distance constraint, and the constraint conditions are as follows:
L(ui,vi)-L(uj,vj)＞ε (11)
Wherein u i、vi represents the abscissa of the center of the ith feature region in the image, ε is a user-defined parameter, and the features of the same region are considered without satisfying constraint (2), and the two are combined. The combined region is divided into a shadow region and an illumination region by the region gray average value, the extracted meteorite pit shadow region is shown in fig. 5, a binary image of the shadow region is marked as B S (u, v), and the number of the shadow regions is marked as n S.
Step 2: according to the shape characteristics of the shadow region, setting the size of an initial search window, carrying out edge search on the navigation features to obtain detected meteorite pit edges, carrying out ellipse fitting on the detected meteorite pit edges based on the geometric mean square distance minimum principle, realizing the preliminary detection of the navigation features, and obtaining the geometric mean square error GRMSE of each fitted ellipse.
And (3) because the shadow region extracted in the step (1) is elliptical, the bright region is non-elliptical, elliptical fitting is carried out on the shadow region according to the imaging characteristics of the region shape, and the center of the ellipse obtained by fitting is taken as the center of the shadow region and is marked as B C＝(uc,vc). And (2) taking the center of the shadow area as the center of a search window, designing the size of the search window according to the size of the shadow area, and searching all edges of the same meteorite pit on the edge detection result in the step (1). The purpose of the local search is to reduce the calculation amount of global search imaging matching and improve the search efficiency.
The search window is a circular area with the center B C of the shadow area as the center and R as the radius, all edges of the same meteorite pit are searched in the area, the size of the circle is defined as k times of the total number of pixels of the shadow area, and the calculation formula of the search radius R n of the nth shadow area is as follows:
Where k is a custom search parameter, the present invention takes k=4. D A is the number of pixels contained in the nth shaded region.
And carrying out ellipse fitting based on a geometric mean square distance minimum principle on the detected meteorite pit edge, realizing the preliminary detection of navigation features, and obtaining the geometric mean square error GRMSE of each fitted ellipse, wherein the calculation steps are as follows.
The fitting approximate mean square error of the two-dimensional curve is:
q is the number of two-dimensional point concentration points, f is described by a finite parameter, which can be expressed as f (x) ≡phi (α, x), α is a smooth function, and for a particular α= (α 1···αr)T, f can be written as f=Φ α (x), we refer to α as a parameter, and x as a variable.
In ellipse fitting, the minimization problem of equation (4) can be reduced to a generalized eigenvector problem, where equation (4) becomes:
In the middle of Is the covariance matrix and trace represents the trace of the matrix.
The fit constraint is:
In the middle of Is a symmetric non-negative definite matrix, D represents the differentiation of the function.
Ellipse fitting is the computation of discrete pointsAnd/>And solving the generalized eigenvector problem m·c=λn·c, the minimum eigenvalue λ and the corresponding eigenvector c gives a solution to the elliptical condition. λ is the geometric mean square error and c is the coefficient of a quadratic equation representing the optimal ellipse. The ellipse fitting result of this example is shown in fig. 6.
Step 3: aiming at the problem of false detection of the edge of the merle pit in the overlapped area, the number of the public edge points searched by the shadow area is used as feedback information to realize dynamic adjustment of the searching radius, and ellipse fitting is carried out on the characteristic edge again until the obtained geometric mean square error is minimum, so that accurate detection of the feature of the merle pit in the overlapped area is realized.
The dynamic adjustment principle refers to: and (3) adjusting the searching radius R n according to the number of the public edge points searched by the shadow region until the geometric mean square error GRMSE of each fitting ellipse obtained in the step (2) reaches the minimum value, wherein the specific implementation mode is as follows.
Through the ellipse detection in the step 2, most merle pits on the surface of the celestial body can be successfully identified, but the overlapped merle pits can search the edges of adjacent features when carrying out edge search, and the fitting precision is poor. This step dynamically adjusts the search radius R n according to the number of common points found by the two features until GRMSE reaches a minimum.
Recording the common point searched by two overlapped meteorites as P n, constructing an adjustment coefficientAnd dynamically adjusting R n until the optimal solution of the ellipse fitting is found. q is a scaling factor between 0 and 1, and the search radius is adjustedThe ellipse fitting problem is converted into the calculation of alpha, and the following constraint problem is written:
The ellipse corresponding to the minimum value of equation (7) is the optimal ellipse, the ellipse fitting result in this example is shown in fig. 7, and the fitting geometric mean square error of each merle is shown in the following table:
Table 1 merle pit fitting mean square error
* Indicating that there is an overlap of this merle with other merles.
Thus, the detection of the complex topography features of the surface of the required celestial body in the deep space detector navigation system is completed.
While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.
Claims (2)
1. The method for detecting the complex topography characteristics of the surface of the celestial body is characterized by comprising the following steps of: comprises the following steps of the method,
Step 1: image processing is carried out on an image shot by a navigation optical camera, edge and region information is extracted, a maximum stable extremum region is combined by utilizing distance constraint so as to remove repeated detection results of the same region, and a shadow region of the surface navigation feature of the celestial body is extracted according to the gray average value of the region;
Step 2: according to the shape characteristics of the shadow region, setting the size of an initial search window, carrying out edge search on the navigation features to obtain detected meteorite pit edges, carrying out ellipse fitting on the detected meteorite pit edges based on the geometric mean square distance minimum principle, realizing the preliminary detection of the navigation features, and obtaining geometric mean square errors GRMSE of fitting ellipses;
The implementation method of the step 2 is that,
Because the shadow area extracted in the step 1 is elliptical, the bright area is non-elliptical, elliptical fitting is carried out on the shadow area according to the imaging characteristics of the area shape, and the elliptical center obtained by fitting is taken as the center of the shadow area and is marked as B C＝(uc,vc); setting the size of a search window according to the size of the shadow area by taking the center of the shadow area as the center of the search window, and searching all edges of the same meteorite pit on the edge detection result in the step 1; the purpose of local search is to reduce the calculation amount of global search imaging matching and improve the search efficiency;
The search window is a circular area with the center B C of the shadow area as the center and R as the radius, all edges of the same meteorite pit are searched in the area, the size of the circle is defined as k times of the total number of pixels of the shadow area, and the calculation formula of the search radius R n of the nth shadow area is as follows:
Wherein k is a custom search parameter; d A is the number of pixels contained in the nth shadow region;
Carrying out ellipse fitting based on a geometric mean square distance minimum principle on the detected meteorite pit edge to realize the preliminary detection of navigation features, and obtaining geometric mean square errors GRMSE of each fitted ellipse, wherein the calculating steps are as follows;
the fitting approximate mean square error of the two-dimensional curve is:
q is the number of two-dimensional point concentration points, f is described by a finite parameter, denoted as f (x) ≡phi (α, x), α is a smooth function, for a particular α= (α 1···αr)T, f is written as f=phi α (x), α is referred to as a parameter, and x is referred to as a variable;
In ellipse fitting, the minimization problem of equation (4) can be reduced to a generalized eigenvector problem, where equation (4) becomes:
In the middle of Is the covariance matrix, trace represents the trace of the matrix;
the fit constraint is:
In the middle of Is a symmetric non-negative definite matrix, D represents the differentiation of the function;
ellipse fitting is the computation of discrete points And/>Solving the generalized eigenvector problem m.c=λn.c, and giving a solution of the elliptical condition by the minimum eigenvalue λ and the corresponding eigenvector c; λ is the geometric mean square error, c is the coefficient of a quadratic equation representing the optimal ellipse;
The dynamic adjustment principle of the search radius refers to: the searching radius R n is adjusted according to the number of the public edge points searched by the shadow region until the geometric mean square error GRMSE of each fitting ellipse obtained in the step 2 reaches the minimum value, and the specific implementation method is as follows,
After the ellipse preliminary detection in the step2, most merle pits on the surface of the celestial body can be successfully identified, but the overlapped merle pits can search the edges of adjacent features when carrying out edge search, so that the fitting precision is poor; recording the common point searched by two overlapped meteorites as P n, constructing an adjustment coefficientDynamically adjusting R n until an optimal solution of ellipse fitting is found; beta is a scaling factor between 0 and 1, and the search radius/>The ellipse fitting problem is converted into alpha;
Written as constraint problem as shown in equations (7) (8) (9):
The ellipse corresponding to the minimum value is the optimal ellipse in the formula (7), so that the precise detection of the meteorite pit characteristics with the overlapped area is realized;
Step 3: aiming at the problem of false detection of the edge of the meteorite crater in the overlapped area, the number of the common edge points searched by the shadow area is used as feedback information to realize dynamic adjustment of the searching radius, and ellipse fitting is carried out on the characteristic edge again until the obtained geometry is uniform
The square error is minimum, and the precise detection of the merle pit characteristics with the overlapped area is realized.
2. The method for detecting complex topography features on a celestial surface of claim 1, wherein: the implementation method of the step 1 is that,
Step 1.1, edge detection is carried out on the merle pit on the surface of the celestial body in the image of the navigation camera, so that edge information of the image is obtained, and short edges and pseudo edges are removed;
Based on a canny edge detection algorithm, edge detection is carried out on the merle pit on the surface of the celestial body, short edges are removed through a communication area mark, the extracted binary image edge information is marked as B (u, v), the corresponding gradient vector is marked as g (u, v), and u and v respectively represent the horizontal coordinates and the vertical coordinates of edge pixel points in the image; the gradient directions of the false edges detected in the meteorite crater and the true edges of the meteorite crater are opposite, and the elimination of the false edge features is realized through the illumination direction; the included angle between the gradient direction of the meteorite crater edge and the direction of the light source is smaller than 90 degrees, and for the image gradient vector g (u, v) of the edge and the direction vector n of the light source on the image plane, the true edge of the meteorite crater satisfies:
the false edge rejection is realized by using the constraint shown in the formula (1), and the edge information of the image is obtained;
Step 1.2, extracting the maximum stable extremum region of the navigation image, utilizing the distance constraint to realize region merging and removing repeated detection results of the same topographic feature, and extracting the shadow region of the celestial body surface navigation feature according to the region gray average value;
Based on the MSER characteristic of the maximum stable extremum region, performing image processing on the merle on the surface of the celestial body to obtain the region characteristic of the image, and primarily dividing the merle image into a shadow region image and a bright region image; the basic principle of the MSER is that the gray level images are sequentially subjected to binarization processing by taking increasing thresholds, and in all obtained binary images, the area with small change of the connected area is called as the maximum stable extremum area; the center of the MSER feature is marked as L (u, v), the obtained region features are subjected to region combination through the constraint condition shown in the formula (2) by virtue of the distance constraint, so that repeated detection results of the same feature are removed, and a shadow region of the celestial body surface navigation feature is extracted according to the region gray average value;
L(ui,vi)-L(uj,vj)＞ε(2)
Wherein ui, vi represent the abscissa of the center of the ith feature region in the image, ε is a user-defined parameter, and consider the same region feature that does not satisfy constraint (2), combine the two; the region is divided into a shadow region and an illumination region by the region gray average value, the binary image of the shadow region of the extracted navigation feature is marked as B S (u, v), and the number of the shadow regions is marked as n S.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111472200.3A CN114187529B (en) | 2021-12-06 | 2021-12-06 | Small celestial body surface complex terrain feature detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111472200.3A CN114187529B (en) | 2021-12-06 | 2021-12-06 | Small celestial body surface complex terrain feature detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114187529A CN114187529A (en) | 2022-03-15 |
| CN114187529B true CN114187529B (en) | 2024-05-31 |
Family
ID=80542275
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111472200.3A Active CN114187529B (en) | 2021-12-06 | 2021-12-06 | Small celestial body surface complex terrain feature detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114187529B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116665062A (en) * | 2023-07-25 | 2023-08-29 | 山东中科冶金矿山机械有限公司 | Mineral resource monitoring method based on remote sensing image |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006350553A (en) * | 2005-06-14 | 2006-12-28 | Kyoto Univ | Corresponding point retrieval method, mutual location method, three-dimensional image measurement method, corresponding point retrieval device, mutual location device, three-dimensional image measurement device, corresponding point retrieval program and computer-readable recording medium with its program recorded |
| CN102944226A (en) * | 2012-12-03 | 2013-02-27 | 哈尔滨工业大学 | Meteor crater detecting method based on bright and dark area pairing |
| CN109708643A (en) * | 2019-01-14 | 2019-05-03 | 北京理工大学 | Asteroid surface optical navigation road sign evaluates choosing method |
| CN110619368A (en) * | 2019-09-23 | 2019-12-27 | 北京理工大学 | Planet surface navigation feature imaging matching detection method |
| CN111721302A (en) * | 2020-06-24 | 2020-09-29 | 北京理工大学 | Method for recognizing and sensing complex terrain features on surface of irregular asteroid |
-
2021
- 2021-12-06 CN CN202111472200.3A patent/CN114187529B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006350553A (en) * | 2005-06-14 | 2006-12-28 | Kyoto Univ | Corresponding point retrieval method, mutual location method, three-dimensional image measurement method, corresponding point retrieval device, mutual location device, three-dimensional image measurement device, corresponding point retrieval program and computer-readable recording medium with its program recorded |
| CN102944226A (en) * | 2012-12-03 | 2013-02-27 | 哈尔滨工业大学 | Meteor crater detecting method based on bright and dark area pairing |
| CN109708643A (en) * | 2019-01-14 | 2019-05-03 | 北京理工大学 | Asteroid surface optical navigation road sign evaluates choosing method |
| CN110619368A (en) * | 2019-09-23 | 2019-12-27 | 北京理工大学 | Planet surface navigation feature imaging matching detection method |
| CN111721302A (en) * | 2020-06-24 | 2020-09-29 | 北京理工大学 | Method for recognizing and sensing complex terrain features on surface of irregular asteroid |
Non-Patent Citations (2)
| Title |
|---|
| 行星表面陨石坑检测与匹配方法;冯军华等;航空学报;20100925(第09期);全文 * |
| 行星表面非规则陨石坑检测与识别方法;朱圣英等;宇航学报;20130330(第03期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114187529A (en) | 2022-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111028277B (en) | SAR and optical remote sensing image registration method based on pseudo-twin convolution neural network | |
| CN110619368B (en) | Planet surface navigation feature imaging matching detection method | |
| CN107677274B (en) | Unmanned plane independent landing navigation information real-time resolving method based on binocular vision | |
| CN103822616A (en) | Remote-sensing image matching method with combination of characteristic segmentation with topographic inequality constraint | |
| CN111652896B (en) | Method for detecting coarse-fine meteorite crater by inertial navigation assistance | |
| CN104123730A (en) | Method and system for remote-sensing image and laser point cloud registration based on road features | |
| CN107341781B (en) | SAR image correction method based on improved phase consistency feature vector base map matching | |
| CN105046251A (en) | Automatic ortho-rectification method based on remote-sensing image of environmental No.1 satellite | |
| CN103035004B (en) | The method of circular target centralized positioning under a kind of Large visual angle | |
| CN110569861A (en) | Image matching positioning method based on point feature and contour feature fusion | |
| CN111721302B (en) | Method for recognizing and sensing complex terrain features on surface of irregular asteroid | |
| CN114187529B (en) | Small celestial body surface complex terrain feature detection method | |
| CN105547287A (en) | Irregular coelonavigation sight line information extracting method | |
| CN106886988B (en) | Linear target detection method and system based on unmanned aerial vehicle remote sensing | |
| CN113834625A (en) | Aircraft model surface pressure measuring method and system | |
| CN103679740B (en) | ROI (Region of Interest) extraction method of ground target of unmanned aerial vehicle | |
| Yao et al. | Automatic extraction of road markings from mobile laser-point cloud using intensity data | |
| CN113295171B (en) | Monocular vision-based attitude estimation method for rotating rigid body spacecraft | |
| Hu et al. | An approach of semiautomated road extraction from aerial image based on template matching and neural network | |
| CN113947724A (en) | Automatic line icing thickness measuring method based on binocular vision | |
| CN103093241B (en) | Based on the remote sensing image nonuniformity cloud layer method of discrimination of homogeneity process | |
| Zhang et al. | Accurate real-time SLAM based on two-step registration and multimodal loop detection | |
| CN110211148B (en) | Underwater image pre-segmentation method based on target state estimation | |
| CN113838072B (en) | High-dynamic star map image segmentation method | |
| CN111696054B (en) | Rubber dam body detection method based on full-polarization SAR image |
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 |