CN103217163A - Star pattern identification method and star pattern identification device of hybrid redundancy feature - Google Patents
Star pattern identification method and star pattern identification device of hybrid redundancy feature Download PDFInfo
- Publication number
- CN103217163A CN103217163A CN2013101119649A CN201310111964A CN103217163A CN 103217163 A CN103217163 A CN 103217163A CN 2013101119649 A CN2013101119649 A CN 2013101119649A CN 201310111964 A CN201310111964 A CN 201310111964A CN 103217163 A CN103217163 A CN 103217163A
- Authority
- CN
- China
- Prior art keywords
- star
- points
- point
- navigation
- map
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 48
- 238000005259 measurement Methods 0.000 claims abstract description 18
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000010845 search algorithm Methods 0.000 claims description 12
- 238000003909 pattern recognition Methods 0.000 claims description 8
- 238000012567 pattern recognition method Methods 0.000 claims description 7
- 239000013598 vector Substances 0.000 description 16
- 238000010586 diagram Methods 0.000 description 9
- 238000009826 distribution Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 230000007423 decrease Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- PCTMTFRHKVHKIS-BMFZQQSSSA-N (1s,3r,4e,6e,8e,10e,12e,14e,16e,18s,19r,20r,21s,25r,27r,30r,31r,33s,35r,37s,38r)-3-[(2r,3s,4s,5s,6r)-4-amino-3,5-dihydroxy-6-methyloxan-2-yl]oxy-19,25,27,30,31,33,35,37-octahydroxy-18,20,21-trimethyl-23-oxo-22,39-dioxabicyclo[33.3.1]nonatriaconta-4,6,8,10 Chemical compound C1C=C2C[C@@H](OS(O)(=O)=O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2.O[C@H]1[C@@H](N)[C@H](O)[C@@H](C)O[C@H]1O[C@H]1/C=C/C=C/C=C/C=C/C=C/C=C/C=C/[C@H](C)[C@@H](O)[C@@H](C)[C@H](C)OC(=O)C[C@H](O)C[C@H](O)CC[C@@H](O)[C@H](O)C[C@H](O)C[C@](O)(C[C@H](O)[C@H]2C(O)=O)O[C@H]2C1 PCTMTFRHKVHKIS-BMFZQQSSSA-N 0.000 description 1
- 208000010392 Bone Fractures Diseases 0.000 description 1
- 206010017076 Fracture Diseases 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- JJWKPURADFRFRB-UHFFFAOYSA-N carbonyl sulfide Chemical compound O=C=S JJWKPURADFRFRB-UHFFFAOYSA-N 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Images
Landscapes
- Navigation (AREA)
- Complex Calculations (AREA)
Abstract
The invention discloses a star pattern identification method of a hybrid redundancy feature. The star pattern identification method comprises the following steps of: based on the geometric position relation of the navigation star points, establishing a hybrid redundancy pattern feature of a navigation reference star point, and then, creating a binary feature database of the navigation reference star point; based on the geometric position relation of a star map and the star points, calculating the hybrid feature of the star map reference star point; according to the hybrid feature of the star map reference star point, selecting candidate star points corresponding to the star map reference star point from the binary feature database by using similar score measurement; by a maximum matching pair algorithm, selecting a matched star point for the star map reference star point from all the candidate star points. The invention also discloses a star pattern identification device of the hybrid redundancy feature. With the method and the device, high accurate identification rate can be realized under the condition of high star point position noise and star magnitude noise, moreover, acquisition of the matched star point is guaranteed when few accompanying star points exist in the radius of the navigation star point pattern.
Description
Technical Field
The invention relates to the field of star map identification, in particular to a star pattern identification method and device with mixed redundant characteristics.
Background
At present, a star sensor is generally used for spacecraft attitude measurement, and the working principle is as follows: the star map is collected by using a star sensor, and the obtained star map is compared with a database storing known star point positions to complete star map identification, so that the posture with the precision reaching several angular seconds is obtained. In the process, star pattern recognition is a key technology in spacecraft attitude measurement, a pattern star pattern recognition algorithm in the star pattern recognition algorithm is an important recognition algorithm, and the establishment of a star point annular pattern is an important development direction of the pattern star pattern recognition algorithm. The star point annular mode is as follows: selecting a navigation star point as an annular center, and taking all star points in a specified radius around the navigation star point as accompanying star points; calculating the angular distance between the navigation star point and the satellite points to obtain the positions of the satellite points in the annular band, and establishing a lookup table for expressing the radial mode characteristics; and screening the lookup table to obtain the final matching star points of the navigation star points.
However, the above star map identification algorithm in the star point annular mode does not fully consider the influence of the star point position noise on the position of the satellite points in the annulus and the influence of the satellite point loss caused by the noise of the star points and the like, so that the correct identification rate of the star map identification algorithm is obviously reduced under the conditions of large star point position noise and the noise of the star points and the like; in addition, when the number of the satellite points in the radius of the navigation satellite point mode is small, the established satellite point mode discrimination is reduced, redundant satellite points are increased, and finally matched satellite points cannot be obtained.
Disclosure of Invention
In view of the above, the present invention provides a star pattern recognition method and device with hybrid redundancy features, which can obtain a high correct recognition rate under the conditions of large star point position noise and star and other noises, and ensure that matched star points are obtained when there are few satellite points in the navigation star point pattern radius.
In order to achieve the purpose, the technical scheme of the invention is realized as follows:
the invention provides a star pattern recognition method with mixed redundancy characteristics, which comprises the following steps:
establishing a hybrid redundancy mode characteristic of a navigation reference star point by using a geometric position relation of the navigation star point, and then establishing a binary characteristic database of the navigation reference star point;
calculating the mixing characteristics of star points of the star map reference by using the geometric position relationship between the star points of the star map;
selecting candidate star points corresponding to the star map reference star points from a binary characteristic database by using similarity score measurement according to the mixed characteristics of the star map reference star points;
and selecting the matched star points of the star map reference star points from all the candidate star points by using a maximum matching pair algorithm.
In the above solution, before the establishing of the hybrid redundant mode feature of the navigation reference star point by using the geometric position relationship of the navigation star point and then the creating of the binary feature database of the navigation reference star point, the method further includes: establishing redundant radial characteristics and redundant accompanying characteristics of the navigation reference star points by using the all-celestial coordinate;
or performing mode radius redundancy on the navigation reference star points by using the all-celestial coordinates, and then establishing redundant radial features and redundant accompanying features of the navigation reference star points.
In the above scheme, the establishing of the redundant radial features of the navigation reference star point includes:
selecting star points meeting the equal-star threshold value one by one as navigation reference star points;
calculating the star diagonal distance between the navigation reference star point and the navigation satellite points, and calculating the discrete value of the star diagonal distance and the coding value of the star diagonal distance;
and establishing redundant radial characteristics between the navigation reference star point and the navigation satellite points by using the discrete value of the star diagonal distance and the encoding value of the star diagonal distance.
In the above scheme, the establishing of the redundant satellite characteristic of the navigation reference star point includes:
calculating the star diagonal distance between navigation satellite points with the star equal to or less than the threshold value, and calculating the discrete value of the star diagonal distance and the encoding value of the star diagonal distance;
and establishing the redundant adjoint characteristic of the navigation reference star point by using the discrete value of the star diagonal distance and the encoding value of the star diagonal distance.
In the above scheme, the calculating the mixed features of the star points of the star map reference by using the geometric position relationship between the star points of the star map includes: and calculating the radial characteristic and the accompanying characteristic of the star points in the star map by using the geometric position relation between the centroids of the star points in the star map, and taking the radial characteristic and the accompanying characteristic as the mixed characteristic of the star points in the star map.
In the foregoing solution, the selecting a candidate star point corresponding to the star map reference star point from a binary feature database by using the similarity score metric includes: and comparing the similarity of the star map reference star point mixed features with the mixed redundant mode features in the navigation reference star point binary feature database one by using the similarity score measurement, and selecting one or more candidate star points corresponding to each star map reference star point.
In the above scheme, the selecting a matching star point of the star map reference star point from all candidate star points by using the maximum matching pair algorithm includes:
calculating candidate matching star pairs by using a maximum matching pair search algorithm, and determining an initial search path;
and searching the maximum matching pair, searching the candidate star points according to the searching direction from the bright star point to the dark star point to obtain the maximum matching pair, dynamically calculating the candidate matching star pair when the searching is broken in advance, and performing ambiguity check when the searching path is ambiguous to determine the only searching path.
The invention also provides a star pattern recognition device with mixed redundancy characteristics, which comprises: the system comprises a database module, a reference star point feature calculation module and a star point matching module; wherein,
the database module is used for establishing the mixed redundancy mode characteristics of the navigation reference star points by utilizing the geometric position relationship of the navigation star points, then establishing a binary characteristic database of the navigation reference star points and receiving the query of the star point matching module on the binary characteristic database;
the star point matching module is used for matching star points of the star map to obtain a geometric position relation between star points of the star map;
the star point matching module is used for selecting candidate star points corresponding to the star map reference star points from a binary characteristic database of the database module by using similarity score measurement according to the mixed characteristics of the star map reference star points sent by the reference star point characteristic calculation module; and selecting the matched star points of the reference star points of each star map from all the candidate star points by using a maximum matching pair algorithm.
The star pattern recognition method and device of the hybrid redundant feature provided by the invention can establish the hybrid redundant pattern feature of the navigation reference star point by utilizing the geometric position relationship of the navigation star point, and further establish a binary feature database of the navigation reference star point; calculating the mixing characteristics of star points of the star map reference by using the geometric position relationship between the star points of the star map; selecting candidate star points corresponding to the star map reference star points from a binary characteristic database by using similarity score measurement according to the mixed characteristics of the star map reference star points; and selecting the matched star points of the reference star points of each star map from all the candidate star points by using a maximum matching pair algorithm. Therefore, the redundant radial mode characteristic of the navigation reference star point is established, and the influence of star point position noise, star noise and the like on the star point mode is fully considered, so that the robustness on the star point position noise and the star noise is better; and the redundant satellite mode of the navigation reference star point is established by utilizing the geometric position relation between the navigation satellite points, the information between the star points is utilized more comprehensively, and the separability of the mode characteristic is increased.
In addition, a hybrid redundancy pattern is constructed by establishing a redundant radial pattern and a redundant companion pattern using a similar discrete coding strategy and storing the redundant patterns using a binary bit string. The hybrid redundancy mode fully utilizes the geometric position information among the star points and can more fully describe the geometric distribution among the star points. The star pattern recognition algorithm provided by the invention not only can have high correct recognition rate under the condition that the number of star points in a field of view is large, but also can well complete recognition under the condition that only a small number of star points in the field of view are detected.
Furthermore, the similarity score measurement is carried out on the similarity of the mixed redundant pattern features by using the pattern features expressed by combining binary bit strings and using the similarity score measurement of shift operation and logic AND operation, so that the comparison speed of pattern feature vectors is increased. The method comprises the steps of searching candidate matching star pairs and then searching the maximum matching pair through a maximum matching pair search algorithm, searching and obtaining the final identification star points of star map reference star points from the candidate matching star points, effectively improving the speed of a star map identification algorithm, particularly accelerating the speed of obtaining the final matching star points through screening from a large number of candidate star points, and quickly finishing star map identification.
Drawings
FIG. 1 is a schematic flow chart of a star pattern recognition method with hybrid redundancy features according to the present invention;
FIG. 2 is a schematic diagram of the redundant radial feature creation of a navigation reference star point according to the present invention;
FIG. 3 is a schematic diagram of a redundant radial feature creation process of the present invention;
FIG. 4 is a schematic diagram of the redundant satellite characteristic establishment of the navigation reference star point according to the present invention;
FIG. 5 is a schematic diagram of creating a star point relationship for radius redundancy;
FIG. 6 is a schematic diagram of a binary feature database structure of navigation reference star points;
FIG. 7 is a schematic diagram of the searching process of the maximum matching pair algorithm;
FIG. 8 is a schematic structural diagram of a star pattern recognition device incorporating redundancy features of the present invention;
FIG. 9 is a table of parameters for a simulation experiment;
FIG. 10 shows the correct recognition rate of the present invention under different noise of star point positions;
FIG. 11 shows the correct recognition rate of the present invention under different kinds of star noises;
FIG. 12 shows the correct recognition rate for different numbers of stars in the field of view.
Detailed Description
The basic idea of the invention is: establishing a hybrid redundancy mode characteristic of a navigation reference star point by using a geometric position relation of the navigation star point, and then establishing a binary characteristic database of the navigation reference star point; calculating the mixing characteristics of star points of the star map reference by using the geometric position relationship between the star points of the star map; selecting candidate star points corresponding to the star map reference star points from a binary characteristic database by using similarity score measurement according to the mixed characteristics of the star map reference star points; and selecting the matched star points of the star map reference star points from all the candidate star points by using a maximum matching pair algorithm.
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
The star pattern recognition method with mixed redundancy characteristics, as shown in fig. 1, comprises the following steps:
step 101: and establishing the hybrid redundant mode characteristics of the navigation reference star points by using the geometric position relationship of the navigation star points, and then establishing a binary characteristic database of the navigation reference star points.
Step 102: and calculating the mixing characteristics of the star points of the star map reference by using the geometric position relationship among the star points of the star map.
Step 103: and selecting candidate star points corresponding to the star map reference star points from a binary characteristic database by using similarity score measurement according to the mixed characteristics of the star map reference star points.
Step 104: and selecting the matched star points of the star map reference star points from all the candidate star points by using a maximum matching pair algorithm.
Preferably, the method for calculating the hybrid redundant mode feature of the navigation reference star point includes: and establishing redundant radial characteristics and redundant accompanying characteristics of the navigation reference star points by using the all-celestial coordinates.
Preferably, the establishing of the redundant radial features of the navigation reference star point includes: when the position noise of the star point is large, the traditional radial mode cannot accurately describe the radial distribution between the navigation reference star point and the navigation satellite point, and the redundant radial mode is established by performing redundant storage on the traditional radial mode, so that the radial distribution between the star points is more accurately described; the method specifically comprises the following steps:
for any one navigation reference star point P, a redundant radial mode of the navigation reference star point P is established by using all navigation satellite points of the navigation reference star point P according to the following steps:
selecting star points such as stars and the like meeting a threshold value one by one as navigation reference star points; the star-like threshold is a value preset according to actual conditions, and can be 6;
calculating the star diagonal distance between the navigation reference star point and the navigation satellite points, and calculating the discrete value of the star diagonal distance and the coding value of the star diagonal distance;
the discrete value of the star diagonal distance is calculated by using a formula
Calculation of, whereinpiRepresenting the angular distance between the navigation reference star point p and the navigation satellite point i, and e is the angular distance subdivision of the annular band;
the encoding value for calculating the star diagonal distance is a formula
And (c) calculating, wherein,
is the star diagonal of the navigation reference star point and the navigation satellite pointDiscrete value of distance, VpiIs a code value of the angular distance of the star, which represents the annulus index of the navigation satellite points in FIG. 2]Representing a rounding operation.
And establishing redundant radial characteristics between the navigation reference star point and the navigation satellite points by using the discrete value of the star diagonal distance and the encoding value of the star diagonal distance.
The establishing of the redundant radial feature between the navigation reference star point and the navigation satellite point, for example, the establishing of the redundant radial feature between the navigation reference star point p and the navigation satellite point j, as shown in fig. 3, includes:
comparing discrete values of the star-to-angular distances
Whether or not it is less than the coded value (V) of the star diagonal distancepi) Add inner loop redundancy threshold (σ)low) If the sum is less than the sum, the Vth of the redundant radial feature of the navigation reference star point is addedpiBit, and the V thpiSubtracting one bit position to be one;
if not, comparing the discrete values of the star diagonal distances
Code value V whether greater than star diagonalpiPlus outer-loop redundancy threshold (sigma)high) If the sum is larger than the preset value, the Vth radial characteristic of the redundant radial characteristic of the navigation reference star point is addedpiBit, and the V thpiSetting a bit position as one;
if not, the Vth of the redundant radial feature of the navigation reference star point is compared with the Vth of the redundant radial feature of the navigation reference star pointpiThe bit position is one.
Wherein, the sigmalow∈(0，0.5]，σhighE [0.5, 1). The redundant radial pattern vector of the navigation reference star point p comprises NmaxAnd a bit, if the j th bit position is 1, indicating that the j th zone has the navigation satellite, and if not, setting 0 to indicate that the navigation satellite is not provided.Wherein, the N ismaxThe total bit number of the redundant radial pattern vector representing the navigation reference star point, i.e. the redundant radial pattern vector of each navigation reference star point has NmaxBit, NmaxMode radius/annulus fine division. In order to make efficient use of memory space, N is the preferred embodiment of the present inventionmax＝6.65/0.03≈222，NmaxTake 224, see the parameters in FIG. 9.
For example, fig. 2(a) does not satisfy the redundancy condition, and a non-redundant radial pattern is established, only bit position 3; fig. 2(b) shows that the star point a satisfies the inner ring redundancy condition, and therefore the 2 nd and 3 rd bit positions of the feature vector are also set to 1, so as to establish the radial pattern of the inner ring redundancy, and fig. 2(c) shows that the star point a satisfies the outer ring redundancy condition, and therefore the 3 rd and 4 th bit positions of the feature vector are also set to 1, so as to establish the radial pattern of the outer ring redundancy. And establishing a redundant radial mode vector of the navigation reference star point according to the method, and increasing the robustness of the mode characteristics to the star point position noise.
Preferably, the establishing of the redundant satellite characteristic of the navigation reference star point includes: calculating the star diagonal distance, the discrete star diagonal distance and the encoding star diagonal distance between navigation satellite points; and according to the flow shown in fig. 3, establishing the redundant adjoint feature of the navigation reference star point by using the discrete value of the star diagonal distance and the coded value of the star diagonal distance.
For example, as shown in FIG. 4, the P star is a navigation satellite, the ABCD star is a navigation satellite selected to establish a redundant satellite pattern with a brightness greater than a threshold, and the E star is dark and not selected. In fig. 4(b), the discrete value of the AB star diagonal is 24.93, and the corresponding encoded angular distance value of 24 (rounded) satisfies the outer-loop redundancy, so the 24 th and 25 th bit positions are set to 1 at the same time; the discrete value of the BD star diagonal distance is 57.04, the corresponding coding angular distance value is 57, inner ring redundancy is met, and the 56 th bit position and the 57 th bit position are simultaneously set to be 1; and the AD star pair does not satisfy inner ring redundancy or outer ring redundancy, and only has bit position 1 at the 17 th position.
Wherein the calculating the star diagonal distance between the navigation satellite points comprises: selecting navigation reference star point P aroundAt most NaCalculating an angular distance upsilon between a navigation satellite point i and a navigation satellite point j, wherein the brightest navigation satellite points are associated with the navigation satellite points, the navigation satellite points and the like are smaller than a threshold valueij，0＜i＜j＜Na. The threshold is a value preset according to actual conditions, and can be 6.5. The discrete star diagonal distance and the encoding star diagonal distance comprise: according to the formula
Performing discrete star diagonal distance and using formula
The star diagonal is encoded and the redundant companion patterns are stored using the same binary bit string as the redundant radial patterns.
Wherein upsilon isijIs the star diagonal distance between the satellite point i and the satellite point j,
is a discrete value of the star diagonal distance between the satellite point i and the satellite point j, VijIs the encoded value of the star diagonal distance, which represents the index of the companion diagonal distance in the redundant companion pattern, PR is the pattern radius, NmaxIs the maximum index number [, ]]Is a rounding operation. According to the discrete value of the star diagonal distance
Sum-star diagonal code value VijThe redundant companion patterns for companion stars are constructed in a similar process to that shown in figure 3. Fig. 4(b) shows PR 6.65 °, Nmax224, navigation is accompanied by discrete values of star diagonal distance between the star points; FIG. 4(c) depicts the equation when σlow＝0.3，σhighAt 0.7, the result of the redundant companion mode vector is constructed using the star diagonal discrete values and the encoded values.
Therefore, the traditional mode class algorithm does not describe the mutual positions of navigation satellite points, and simultaneously does not fully consider the influence of satellite point loss caused by noise such as satellite points and the like, and the redundant satellite characteristics are constructed according to the geometric positions of the navigation satellite points, so that the separability of the satellite point mode characteristics is enhanced, and the overlapping degree of the navigation reference satellite point mode characteristics in the characteristic database is reduced; meanwhile, the redundancy accompanying characteristics established by the invention do not need to select positioning star points to establish a uniform coordinate system, and the star point mode has certain position redundancy, so that the algorithm can effectively inhibit star equal noise and star point position noise, and the robustness of the star map identification algorithm to the star equal noise and the star point position noise is improved.
Preferably, the redundant radial features and redundant satellite features for establishing the navigation reference star points may further be: and performing mode radius redundancy on the navigation reference star points, and then establishing redundant radial features and redundant accompanying features of the navigation reference star points. The operation of mode radius redundancy on the navigation reference star points is only one choice, and a mixed redundancy characteristic can be obtained by directly calculating each navigation reference star point in the database; and for part of navigation reference star points, after the common mixed redundancy feature is established, the mixed redundancy feature under the condition of radius redundancy needs to be established. Therefore, in fig. 6, the jth navigation reference star point has some more pattern features than the ith navigation reference star point, and the jth navigation reference star point not only includes a redundant radial pattern and a redundant satellite pattern in a normal case, but also includes a redundant radial pattern and a redundant satellite pattern in a radius redundancy case.
The mode radius redundancy of the navigation reference star point is as follows: if navigation satellite points around a navigation reference satellite point are generally darker, the reference satellite point can be an isolated satellite point when the detection capability of a star sensor is weaker, and in order to improve the identification rate of a star map identification algorithm under the condition, mode radius redundancy is performed on navigation reference satellite points such as navigation satellite points and stars which are all larger than a set threshold value, a hybrid redundancy mode is created, for example, as shown in fig. 5, (a) satellite points within one-time mode radius are dim and easy to be lost collectively, and (b) satellite points outside one-time mode radius within two-time mode radius are selected as satellite points, and the hybrid redundancy mode is established. Wherein the set threshold is set according to the minimum sensitivity capability of the actual star sensor, and in the embodiment of the present invention, may be 6, see fig. 9.
Preferably, the creating a binary feature database of the navigation reference star points includes: and numbering the navigation reference star points as indexes, and recording the star and the like, right ascension and declination coordinates, redundant radial features and redundant accompanying features of the navigation reference star points. Since the redundant radial patterns and the redundant concomitant patterns established for the navigation reference star points by the invention are a set of different types of pattern features, the redundant radial patterns and the redundant concomitant patterns are created according to similar discrete coding and redundancy strategies and are stored by using binary bit strings; therefore, the method can be combined into a hybrid redundancy mode, and then shift operation and logic AND operation can be used for quickly identifying star points, so that all information of a redundancy radial mode and a redundancy concomitant mode can be comprehensively utilized.
The invention adopts a similar strategy to establish the redundant mode of the star points, so that two different redundant modes can be combined into a mixed redundant mode, when the number of the star points in the star map is more, the redundant radial mode has larger influence on a matching result, and when the number of the star points in the star map is less, the influence of the redundant concomitant mode is more obvious, therefore, the mixed redundant mode fully utilizes the geometric position information between the star points and can more completely describe the geometric distribution between the star points. The star map identification algorithm provided by the invention not only can have high correct identification rate under the condition that the number of star points in a field of view is large, but also can well complete identification under the condition that only a small number of star points in the field of view are detected, and has better adaptability compared with the traditional mode star map identification algorithm. Because there are many navigation reference star points in the celestial sphere, the hybrid redundant pattern features of the navigation reference star points need a large storage space for storage, and therefore, the star map recognition algorithm needs a good data storage structure to reduce the spatial complexity of the pattern feature database. The invention uses the binary bit string with smaller space complexity to store the mixed redundancy mode characteristics of the navigation reference star point, thereby effectively reducing the storage space required by the characteristic database. For example, fig. 6 is a database structure of navigation reference star point pattern features, a storage structure of navigation reference star point basic information and its mixed redundant pattern features, and n is the total number of navigation reference star points. Navigation satellite points around the navigation reference satellite point j are dim and can cause the reference satellite point to be an isolated satellite point, so that a hybrid redundancy mode under the condition of mode radius redundancy needs to be established besides a redundancy radial mode and a redundancy satellite mode. The redundant features of the invention are all stored by adopting binary bit strings, on one hand, the storage space is saved, on the other hand, the unified comparison of the redundant mode feature similarity can be realized by adopting the shift and logic and operation with high operation speed, and the star atlas identification speed is accelerated.
Preferably, the calculating the mixed feature of the star map reference star points by using the geometric position relationship between the star map star points includes: and establishing the radial characteristic and the accompanying characteristic of the star map reference star point as the mixed characteristic of the star map reference star point by using the geometric position relation between the centroids of the star points in the star map. Preferably, the method for determining the centroid of the star points in the star map comprises the following steps: the star sensor collects a star map, extracts star points in the star map and calculates the mass center of the star points; the extracting of star points in the star map and the calculating of the centroid of the star points are in the prior art, and are not described herein.
Preferably, the selecting the candidate star point corresponding to the star atlas reference star point from the binary feature database by using the similarity score metric includes: and comparing the similarity of the mixed features of the star map reference star points with the mixed redundant mode features in the navigation reference star point binary feature database one by using the similarity score measurement, and selecting one or more candidate star points corresponding to each star map reference star point.
The star table is used for creating a navigation reference star point mixed redundant mode characteristic database in an off-line mode, star images are collected through a star sensor, the mixed characteristics of star image reference star points are obtained on line, and a star point mode characteristic matching algorithm compares the mixed redundant mode characteristics in the binary characteristic database of the navigation reference star points with the mixed mode characteristics of the star image star points to achieve the unique matching of the star image star points and the navigation reference star points. However, the unique matching relationship cannot be directly obtained, and the star map star point mixed mode feature is similar to the mixed redundant feature of several or dozens of navigation reference star points, so that redundant matching occurs. These several or dozens of navigation reference star points are referred to as candidate star points with respect to the star map reference star points, which are referred to as candidate star points for short.
The similarity score metric may be calculated using the following formula:
wherein,
patternr[i]and patternc[j]Respectively representing the value of the ith bit of the radial pattern of the star point of the star map and the value of the jth bit of the adjoint pattern,
and
respectively representing the value of the ith bit of the redundant radial mode and the value of the jth bit of the redundant adjoint mode of the navigation reference star point p;scorep represents the similarity score of the star map reference star point and the navigation reference star point p, M is the total number of valid bits (the value in a bit is 1) in a radial mode of the star map star point, N is the total number of valid bits (the value in the bit is 1) in an accompanying mode of the star map star point, R is a weight coefficient of the similarity score of the accompanying mode, and through multiple times of experimental verification, R is 1.0.
The one or more candidate star points corresponding to the reference star point of each star map are selected as follows: and selecting the navigation reference star higher than the similarity percentage threshold value as a candidate star point. Since the similarity score metric of the pattern vector uses shift operation and logical and operation with fast calculation speed, it is 30% faster than the pattern class algorithm using the traditional lookup table in the screening to obtain the candidate star point.
Preferably, the maximum matching pair algorithm is as follows: after comparing the star point pattern characteristics through similarity score measurement, some ambiguous matching star points exist, and in order to obtain the only matching star points, the invention provides a maximum matching pair search algorithm to search the final matching star points from the candidate matching star points. According to the principle that the matching star points can form the maximum number of matching star pairs, the matching star points are quickly searched from a large number of candidate star points by searching from bright star points to dark star points and searching for the maximum matching pair after searching for the candidate matching star pairs.
Specifically, as shown in fig. 7, candidate matching star pairs of the star map star pairs are calculated, and an initial search path is determined; searching the maximum matching pair, and searching the candidate star points according to the searching direction from the bright star point to the dark star point to obtain the maximum matching pair; when the search is broken in advance, dynamically calculating candidate matching star pairs; and if the search path is ambiguous, performing ambiguity check to determine the unique search path.
Wherein the candidate matching star pairs of the star map star pairs are calculated as follows: sequentially calculating star map angular distances between star points of two star maps, calculating celestial sphere angular distances between certain two candidate star points of the current two star map star points, and if the difference value between the star map angular distances and the celestial sphere angular distances is smaller than a threshold value, matching the celestial sphere angular distances of the candidate star points with the star map angular distances, wherein the two candidate star points form a candidate matching star pair of the current two star map star points;
and traversing all the star pairs which can be formed by the candidate matching star points of the two selected star points of the star map, and forming the candidate matching star pair of the star map star pair by the star pair with the celestial sphere angular distance and the star map angular distance.
Wherein the determining the initial search path is: and the star points of the star map arranged in descending order according to the brightness are the initial search path. For example, the initial search path reflects a curved arrow in the left diagram of fig. 7(b) in the star diagram, and the initial search path reflects a search from the AB star pair, through BC, CD, and to the DE star pair in the search algorithm.
The search is broken ahead of time as: if the current candidate matching star pair does not have a common star point with the set of next candidate matching star pairs, searching for a break.
The dynamic calculation candidate matching star pair is as follows: selecting the next nearest neighbor star map star points at the end of the current maximum matching pair and the brightest star points in the unmatched star map reference star point set, calculating the star map star diagonal distance of the selected two star points and the star-to-celestial angular distance formed by the corresponding candidate star points, and determining the candidate matching star pairs of the selected two star points, thereby dynamically generating a new search path.
And ambiguity occurs in the search path, ambiguity check is carried out, and the only search path is determined as follows: if at least two common star points appear in the current candidate matching star pair and the set of the next candidate matching star pair, search path ambiguity occurs, and ambiguity checking is started; and if the two ambiguous star points are matched with the second nearest star point at the tail end of the current maximum matching pair, selecting the current candidate matching star pair as a correct matching star pair.
In fig. 7, the reference star points are arranged in descending order of brightness, and the area of the star point in the figure represents the brightness of the star point, and the area of the star point represents that the brightness of the star point is high, so the star point in the star map a is the brightest star point in the star map. The maximum matching pair search algorithm searches according to the sequence from bright stars to dark stars, the influence of noise such as stars is reduced to the maximum extent, and the consistency of the identification flow is ensured. Compared with the traditional search algorithm for comparing the star points by the star points and the diagonal distances by the diagonal distances, the maximum matching pair search algorithm searches candidate matching star pairs first and then carries out maximum matching pair search, so that the comparison times of the star diagonal distances are greatly reduced, and the operation efficiency of the algorithm is improved. And finding the maximum matching pair from the star map as the final identification result of the star point of the current star map, such as the maximum matching pair in fig. 7 (e).
In fig. 7(d), the candidate matching star pairs of BC and CD do not have the same candidate star point with respect to the C star point, so that the search is broken, and in order to obtain a larger matching pair, the invention selects a dynamically updated search path to generate the candidate matching star pair of the BD star pair, so that the algorithm continues to search forward, and the search is prevented from terminating early. Fig. 7(e) shows the ambiguity checking process of the search algorithm, which is to check whether the celestial angular distance between the ambiguous star and the next nearest star point at the end of the currently identified stream matches the asterogram angular distance between the corresponding asterogram star points. The celestial angular distances of the celestial star pairs (492, 12) and (492, 862) match the asterisk angular distances between the asterisk star points D, E, thereby introducing search path ambiguity. The angular distance matching star pair (492, 862) is selected as the next search direction by checking whether the celestial distance between the star pair (365, 12) and (365, 862) is close to the star distance of the star pair BD.
The maximum matching pair search algorithm calculates candidate matching star pairs firstly and then carries out maximum matching pair search, the angular distance comparison times are effectively controlled, and compared with the traditional search algorithm for gradually comparing the star diagonal distances, the method can quickly search from a plurality of candidate star points to obtain the final matching star points of the star map star points.
As shown in fig. 8, the star map recognition device for redundant pattern features of the present invention includes: the system comprises a database module 81, a reference star point feature calculation module 82 and a star point matching module 83; wherein,
the database module 81 is used for establishing the hybrid redundancy mode characteristics of the navigation reference star points by using the geometric position relationship of the navigation star points, then establishing a binary characteristic database of the navigation reference star points, and receiving the query of the star point matching module 83 on the binary characteristic database;
the reference star point feature calculation module 82 is used for positioning star points in a star map, calculating the mixed features of the star map reference star points by using the geometric position relationship between the star map star points, and sending the mixed features of the star map reference star points to the star point matching module 83;
a star point matching module 83, configured to select candidate star points corresponding to the star map reference star points from the binary feature database of the database module 81 by using the mixed features of the star map reference star points sent by the reference star point feature calculation module 82 and using similarity score measurement; and selecting the matched star points of the star map reference star points from all the candidate star points by using a maximum matching pair algorithm.
The database module 81 is specifically configured to establish redundant radial features and redundant satellite features of the navigation reference star points by using the all-celestial coordinates.
The database module 81 is specifically configured to establish, for any one navigation reference star point P, a redundant radial mode of the navigation reference star point P using all navigation satellite points thereof according to the following three steps:
1) calculating the star diagonal distance between the navigation reference star point and the navigation satellite point;
2) discrete encoding of star diagonal distances, calculating discrete values of the star diagonal distances and encoding values of the star diagonal distances;
the discrete value of the star diagonal distance is calculated by using a formula
Calculation of, whereinpiRepresenting the angular distance between the navigation reference star point p and the navigation satellite point i, and e is the angular distance subdivision of the annular band;
the encoding value for calculating the star diagonal distance is a formula
And (c) calculating, wherein,
is a discrete value of the star diagonal distance between the navigation reference star point and the navigation satellite point, VpiIs a code value of the angular distance of the star, which represents the annulus index of the navigation satellite points in FIG. 2]Representing a rounding operation.
3) And establishing redundant radial characteristics of the navigation reference star points.
The establishing of the redundant radial features of the navigation reference star points may include: establishing redundant radial characteristics of navigation reference star points one by one;
for example, establishing a redundant radial feature between the navigation reference star point p and the navigation satellite star point j, as shown in fig. 3, includes:
comparing discrete values of the star-to-angular distancesWhether or not it is less than the coded value (V) of the star diagonal distancepi) Add inner loop redundancy threshold (σ)low) If the sum is less than the sum, the Vth of the redundant radial feature of the navigation reference star point is addedpiBit, and the V thpiSubtracting one bit position to be one;
if not, comparing the two
Whether or not it is greater than VpiPlus outer-loop redundancy threshold (sigma)high) If the sum is larger than the preset value, the Vth radial characteristic of the redundant radial characteristic of the navigation reference star point is addedpiBit, and the V thpiSetting a bit position as one;
if not, the Vth of the redundant radial feature of the navigation reference star point is compared with the Vth of the redundant radial feature of the navigation reference star pointpiThe bit position is one.
Wherein, the sigmalow∈(0，0.5]，σhighE [0.5, 1). The radial mode vector of the navigation reference star point p contains NmaxAnd a bit, if the j th bit position is 1, indicating that the j th zone has the navigation satellite, and if not, setting 0 to indicate that the navigation satellite is not provided. Wherein, the N ismaxThe total bit number of the redundant radial pattern vector representing the navigation reference star point, i.e. the redundant radial pattern vector of each navigation reference star point has NmaxBit, NmaxMode radius/annulus fine division. In order to make efficient use of memory space, N is the preferred embodiment of the present inventionmax＝6.65/0.03≈222，NmaxGet 224, see parameter in FIG. 9
For example, fig. 2(a) does not satisfy the redundancy condition, and a non-redundant radial pattern is established, only bit position 3; fig. 2(b) shows that the star point a satisfies the inner ring redundancy condition, and therefore the 2 nd and 3 rd bit positions of the feature vector are also set to 1, so as to establish the radial pattern of the inner ring redundancy, and fig. 2(c) shows that the star point a satisfies the outer ring redundancy condition, and therefore the 3 rd and 4 th bit positions of the feature vector are also set to 1, so as to establish the radial pattern of the outer ring redundancy. And establishing a redundant radial mode vector of the navigation reference star point according to the method, and increasing the robustness of the mode characteristics to the star point position noise.
The database module 81 is specifically configured to calculate the star diagonal distances between the navigation satellite points and the discrete encoding star diagonal distances.
The database module 81 is specifically configured to select at most N around the navigation reference star point PaCalculating an angular distance upsilon between a navigation satellite point i and a navigation satellite point j, wherein the brightest navigation satellite points are associated with the navigation satellite points, the navigation satellite points and the like are smaller than a threshold valueij，0＜i＜j＜Na。
The database module 81 is specifically configured to operate according to the formula
Performing discrete star diagonal distance and using formula
Encoding the star diagonal distance, and creating and storing a redundant adjoint mode by adopting a flow similar to that shown in FIG. 3 and a binary bit string same as the redundant radial mode; wherein upsilon isijIs the star diagonal distance between the satellite point i and the satellite point j,
is a discrete value of the star diagonal distance between the satellite point i and the satellite point j, VijIs the encoded value of the star diagonal distance, which represents the index of the companion diagonal distance in the redundant companion pattern, PR is the pattern radius, NmaxIs the maximum index number [, ]]Is a rounding operation. According to the discrete value of the star diagonal distance
Sum-star diagonal code value VijThe redundant companion patterns for companion stars are constructed in a similar process to that shown in figure 3. Fig. 4(b) shows PR 6.65 °, Nmax224, navigation is accompanied by discrete values of star diagonal distance between the star points; FIG. 4(c) depicts the equation when σlow＝0.3，σhighAt 0.7, the result of the redundant companion mode vector is constructed using the star diagonal discrete values and the encoded values.
The database module 81 is specifically configured to perform mode radius redundancy on the navigation reference star points, and then establish redundant radial features and redundant accompanying features of the navigation reference star points.
The database module 81 is specifically configured to, if navigation satellite points around the navigation reference satellite point are generally dark, the reference satellite point may become an isolated satellite point when the detection capability of the star sensor is weak, and in order to improve the recognition rate of the star map recognition algorithm under such a condition, we perform mode radius redundancy on the navigation reference satellite point, and create a hybrid redundant satellite point mode. For example, as shown in fig. 5, (a) satellite stars within one mode radius are dim and easily missing collectively, and (b) satellite stars outside one mode radius within two mode radii are selected as satellite stars to establish a star point pattern.
The database module 81 is specifically configured to use the number of the navigation reference star point as an index, and record the star, the right ascension and declination coordinates, the redundant radial feature and the redundant accompanying feature of the navigation reference star point.
The reference star point feature calculation module 82 is specifically configured to perform star map acquisition, extract star points in a star map, and calculate a centroid of the star points.
The reference star point feature calculation module 82 is specifically configured to establish a radial feature and an accompanying feature of a star map reference star point as a mixed feature of the star map reference star point by using a geometric position relationship between centroids of star points in the star map.
The star point matching module 83 is specifically configured to compare similarity between the star map reference star point mixed features and the mixed redundant pattern features in the navigation reference star point binary feature database one by using similarity score measurement, and select one or more candidate star points corresponding to the star map reference star point.
The star point matching module 83 is specifically configured to perform similarity comparison, and may use the following formula to calculate: wherein,
patternr[i]and patternc[j]Respectively representing the value of the ith bit of the radial pattern of the star point as a reference star point of the star map and the value of the jth bit of the adjoint pattern,
The value of the jth bit of the value and redundant companion patterns;scorep represents the similarity score of the star points of the star map and the navigation reference star point p, M is the total number of valid bits (the value in a bit is 1) in a radial mode of the star points of the star map, N is the total number of valid bits (the value in the bit is 1) in an accompanying mode of the star points of the star map, R is a weight coefficient of the similarity score of the accompanying mode, and through multiple times of experimental verification, R is 1.0.
The star point matching module 83 is specifically configured to select a navigation reference star higher than the similarity percentage threshold as a candidate star point.
The star point matching module 83 is specifically configured to calculate candidate matching star pairs of the star map star pairs, and determine an initial search path; searching forward by a search algorithm to obtain a maximum matching pair; searching for early fracture, dynamically calculating candidate matching star pairs, and changing a search path; and (4) ambiguity occurs in the search path, ambiguity check is carried out, and the unique search path is determined.
The device can be used as a logic module to be installed in the star sensor in the prior art.
In order to evaluate the performance of the algorithm, an SAO star table containing 15935 total star points such as 7.0 stars is adopted, and a computer simulation star map is used for carrying out a simulation experiment, wherein various parameters of the simulation experiment are shown in fig. 9.
Due to the influence of a plurality of factors such as image noise, optical system distortion, sampling quantization truncation errors, errors of a centroid positioning algorithm and the like, the positioning of the star centroid often has larger errors. In order to investigate the inhibition capability of the algorithm of the invention on the star point centroid positioning error, a certain gaussian noise with the mean value of 0 is added to the star point centroid in the simulated star map, and the identification result of the star map under 10000 random attitudes is counted as shown in fig. 10. When centroid Gaussian noise is small (σ)pLess than 0.5 pixel), the algorithm of the invention can reach more than 99.22% of correct recognition rate; when the star point centroid Gaussian noise is at a moderate noise level (sigma)p0.8 pixels), the algorithm of the present invention can correctly identify 98.41% of star maps; when the star point mass center has large Gaussian noise (sigma)p1.6 pixels), the algorithm of the present invention can still achieve 87.88% correct recognition rate.
Different stars have different spectral distributions, the brightness of some stars constantly changes, the star sensor has different spectral responses to light with different frequencies, and the star sensor has the influence of factors such as inherent noise and the like, so that the brightness of star points detected by the star sensor has certain deviation. By usingGaussian noise with the average value of 0 is added to the brightness of the star points to simulate detection errors of star sensors and the like. FIG. 11 shows the correct recognition rate of the algorithm of the present invention under different star-like noises. The accurate star map recognition rate is continuously reduced along with the increase of star point and star equal noise, and when the star equal noise is smaller (sigma)M0.323 stars, etc.), the algorithm of the present invention can correctly identify 99.29% of star maps; in moderate star equal noise σMWhen the number is 0.646 star and the like, the algorithm can obtain 94.86% correct recognition rate; at a large star equal noise sigmaMThe algorithm still has 86.17% correct recognition rate when equal to 0.969 stars, and other pattern star map recognition algorithms have lower recognition rate in this case.
The star sensor works in outer space for a long time and is easily influenced by cosmic rays, so that the device aging and the sensitivity are reduced, the capacity of the star sensor for sensing stars and the like is reduced under the condition that the data updating rate is not changed, and the number of the stars in a view field is reduced; on the other hand, when the visual axis of the star sensor points to a sky area with rare star points, the number of star points in the visual field is also less. The traditional mode star map identification algorithm has higher correct identification rate only when the star number in the view field is more. FIG. 12 shows the correct recognition rate of the algorithm of the present invention for different numbers of stars in the field of view. As can be seen from fig. 12: the rate of correct recognition decreases substantially as the number of stars in the field of view decreases. The correct recognition rate of the algorithm in a multi-star point view field (33 star points) is 99.19%, the correct recognition rate in a general star point number view field (13 star points) is 98.02%, the correct recognition rate in a sparse star point view field (4 star points) is 77.12%, and the correct recognition rate of the traditional mode star-like image recognition algorithm is very low under the condition that the star points are rare (the number of the star points is less than 10).
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention.
Claims (8)
1. A star pattern recognition method for mixed redundant features is characterized by comprising the following steps:
establishing a hybrid redundancy mode characteristic of a navigation reference star point by using a geometric position relation of the navigation star point, and then establishing a binary characteristic database of the navigation reference star point;
calculating the mixing characteristics of star points of the star map reference by using the geometric position relationship between the star points of the star map;
selecting candidate star points corresponding to the star map reference star points from a binary characteristic database by using similarity score measurement according to the mixed characteristics of the star map reference star points;
and selecting the matched star points of the star map reference star points from all the candidate star points by using a maximum matching pair algorithm.
2. The method of claim 1, wherein before the creating the hybrid redundant mode feature of the navigation reference star point by using the geometric position relationship of the navigation star point and then creating the binary feature database of the navigation reference star point, the method further comprises: establishing redundant radial characteristics and redundant accompanying characteristics of the navigation reference star points by using the all-celestial coordinate;
or performing mode radius redundancy on the navigation reference star points by using the all-celestial coordinates, and then establishing redundant radial features and redundant accompanying features of the navigation reference star points.
3. The method of claim 2, wherein the establishing redundant radial features of the navigation reference star points comprises:
selecting star points meeting the equal-star threshold value one by one as navigation reference star points;
calculating the star diagonal distance between the navigation reference star point and the navigation satellite points, and calculating the discrete value of the star diagonal distance and the coding value of the star diagonal distance;
and establishing redundant radial characteristics between the navigation reference star point and the navigation satellite points by using the discrete value of the star diagonal distance and the encoding value of the star diagonal distance.
4. The method of claim 2, wherein establishing redundant satellite features of navigation reference star points comprises:
calculating the star diagonal distance between navigation satellite points with the star equal to or less than the threshold value, and calculating the discrete value of the star diagonal distance and the encoding value of the star diagonal distance;
and establishing the redundant adjoint characteristic of the navigation reference star point by using the discrete value of the star diagonal distance and the encoding value of the star diagonal distance.
5. The method according to claim 1, wherein the calculating the mixed feature of the star map reference star points by using the geometric position relationship between the star map star points comprises: and calculating the radial characteristic and the accompanying characteristic of the star points in the star map by using the geometric position relation between the centroids of the star points in the star map, and taking the radial characteristic and the accompanying characteristic as the mixed characteristic of the star points in the star map.
6. The method of claim 1, wherein selecting candidate star points from a binary feature database corresponding to the reference star point of the star map using the similarity score metric comprises: and comparing the similarity of the star map reference star point mixed features with the mixed redundant mode features in the navigation reference star point binary feature database one by using the similarity score measurement, and selecting one or more candidate star points corresponding to each star map reference star point.
7. The method of claim 1, wherein selecting a matching star point of the star map reference star point from all candidate star points using a maximum matching pair algorithm comprises:
calculating candidate matching star pairs by using a maximum matching pair search algorithm, and determining an initial search path;
and searching the maximum matching pair, searching the candidate star points according to the searching direction from the bright star point to the dark star point to obtain the maximum matching pair, dynamically calculating the candidate matching star pair when the searching is broken in advance, and performing ambiguity check when the searching path is ambiguous to determine the only searching path.
8. A star pattern recognition apparatus incorporating redundant features, the apparatus comprising: the system comprises a database module, a reference star point feature calculation module and a star point matching module; wherein,
the database module is used for establishing the mixed redundancy mode characteristics of the navigation reference star points by utilizing the geometric position relationship of the navigation star points, then establishing a binary characteristic database of the navigation reference star points and receiving the query of the star point matching module on the binary characteristic database;
the star point matching module is used for matching star points of the star map to obtain a geometric position relation between star points of the star map;
the star point matching module is used for selecting candidate star points corresponding to the star map reference star points from a binary characteristic database of the database module by using similarity score measurement according to the mixed characteristics of the star map reference star points sent by the reference star point characteristic calculation module; and selecting the matched star points of the reference star points of each star map from all the candidate star points by using a maximum matching pair algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310111964.9A CN103217163B (en) | 2013-04-02 | 2013-04-02 | Star pattern identification method and star pattern identification device of hybrid redundancy feature |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310111964.9A CN103217163B (en) | 2013-04-02 | 2013-04-02 | Star pattern identification method and star pattern identification device of hybrid redundancy feature |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103217163A true CN103217163A (en) | 2013-07-24 |
| CN103217163B CN103217163B (en) | 2015-07-15 |
Family
ID=48815167
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310111964.9A Active CN103217163B (en) | 2013-04-02 | 2013-04-02 | Star pattern identification method and star pattern identification device of hybrid redundancy feature |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103217163B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104567865A (en) * | 2014-12-29 | 2015-04-29 | 北京控制工程研究所 | Attitude capture method of star sensor under space particle interference condition |
| CN104776845A (en) * | 2015-04-29 | 2015-07-15 | 西安电子科技大学 | Autonomous star recognition method based on combination mode |
| CN108681617A (en) * | 2018-03-29 | 2018-10-19 | 北京空间飞行器总体设计部 | A kind of spacecraft multi-star sensor layout optimization design method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101004349A (en) * | 2007-01-22 | 2007-07-25 | 北京航空航天大学 | Method for recognising not gauged star map |
| CN102840860A (en) * | 2012-08-30 | 2012-12-26 | 北京航空航天大学 | Hybrid particle swarm algorithm-based star graph recognition method |
-
2013
- 2013-04-02 CN CN201310111964.9A patent/CN103217163B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101004349A (en) * | 2007-01-22 | 2007-07-25 | 北京航空航天大学 | Method for recognising not gauged star map |
| CN102840860A (en) * | 2012-08-30 | 2012-12-26 | 北京航空航天大学 | Hybrid particle swarm algorithm-based star graph recognition method |
Non-Patent Citations (3)
| Title |
|---|
| 许士文等: "用字符匹配进行星图识别的导航库存储方法", 《哈尔滨工业大学学报》, vol. 37, no. 6, 30 June 2005 (2005-06-30), pages 819 - 821 * |
| 魏新国等: "利用径向和环向分布特征的星图识别方法", 《光电工程》, vol. 31, no. 8, 31 August 2004 (2004-08-31), pages 4 - 7 * |
| 魏新国等: "基于Log-Polar变换的星图识别方法", 《光学技术》, vol. 32, no. 5, 30 September 2006 (2006-09-30), pages 678 - 681 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104567865A (en) * | 2014-12-29 | 2015-04-29 | 北京控制工程研究所 | Attitude capture method of star sensor under space particle interference condition |
| CN104567865B (en) * | 2014-12-29 | 2017-05-10 | 北京控制工程研究所 | Attitude capture method of star sensor under space particle interference condition |
| CN104776845A (en) * | 2015-04-29 | 2015-07-15 | 西安电子科技大学 | Autonomous star recognition method based on combination mode |
| CN104776845B (en) * | 2015-04-29 | 2017-05-03 | 西安电子科技大学 | Autonomous star recognition method based on combination mode |
| CN108681617A (en) * | 2018-03-29 | 2018-10-19 | 北京空间飞行器总体设计部 | A kind of spacecraft multi-star sensor layout optimization design method |
| CN108681617B (en) * | 2018-03-29 | 2022-07-29 | 北京空间飞行器总体设计部 | Optimization design method for layout of spacecraft multi-star sensor |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103217163B (en) | 2015-07-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11720790B2 (en) | Method of training image deep learning model and device thereof | |
| CN101441082A (en) | Rapid triangle star map recognition method | |
| CN107816987B (en) | Star pattern recognition method based on spider web pattern and convolutional neural network | |
| Wei et al. | A star identification algorithm based on radial and dynamic cyclic features of star pattern | |
| CN103363987B (en) | The method for recognising star map of a kind of many visual fields star sensor | |
| Xu et al. | RPNet: A representation learning-based star identification algorithm | |
| US20160125243A1 (en) | Human body part detection system and human body part detection method | |
| CN102840860B (en) | A kind of method for recognising star map based on Hybrid Particle Swarm | |
| Zhao et al. | Star identification algorithm based on K–L transformation and star walk formation | |
| US11699085B2 (en) | Methods and arrangements to identify activation profile context in training data | |
| CN106595645A (en) | Method for making guide star database based on output accuracy of star sensors | |
| Jiang et al. | Redundant-coded radial and neighbor star pattern identification algorithm | |
| CN110929080A (en) | Optical remote sensing image retrieval method based on attention and generation countermeasure network | |
| CN103217163B (en) | Star pattern identification method and star pattern identification device of hybrid redundancy feature | |
| Fan et al. | A triangle voting algorithm based on double feature constraints for star sensors | |
| CN106372630A (en) | Face direction detection method based on deep learning | |
| CN103954280A (en) | Rapid, high-robustness and autonomous fixed star identification method | |
| Du et al. | Star identification based on radial triangle mapping Matrix | |
| Yang et al. | Robust and efficient star identification algorithm based on 1-D convolutional neural network | |
| Zhu et al. | A robust and fast star identification algorithm based on an ordered set of points pattern | |
| CN105928509A (en) | Triangle recognizing method based on maximum interior angle hash function | |
| CN107941210B (en) | Star map identification method combining neural network technology and triangle algorithm | |
| CN104833359A (en) | Discrete hidden Markov model characteristic sequence model based star chart pattern recognition method | |
| CN109977961B (en) | Binary feature learning method and system based on layered attention mechanism | |
| CN115131580A (en) | Space target small sample identification method based on attention mechanism |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |