A star identification algorithm was developed for a charge-coupled device （CCD） or complementary metal-oxide-semiconductor （CMOS） autonomous star tracker to acquire 3-axis attitude information for a lost-in-space ...A star identification algorithm was developed for a charge-coupled device （CCD） or complementary metal-oxide-semiconductor （CMOS） autonomous star tracker to acquire 3-axis attitude information for a lost-in-space spacecraft. The algorithm took advantage of an efficient on-board database and an original “4- star matching” pattern recognition strategy to achieve fast and reliable star identification. The on-board database was composed of a brightness independent guide star catalog （mission catalog） and a K-vector star pair catalog. The star pattern recognition method involved direct location of star pair candidates and a sim- ple array matching procedure. Tests of the algorithm with a CMOS active pixel sensor （APS） star tracker result in a 99.9% success rate for star identification for lost-in-space 3-axis attitude acquisition when the angular measurement accuracy of the star tracker is at least 0.01°. The brightness independent algorithm requires relatively higher measurement accuracy of the star apparent positions that can be easily achieved by CCD or CMOS sensors along with subpixel centroiding techniques.展开更多
An artificial intelligence enhanced star identification algorithm is proposed for star trackers in lost-inspace mode.A convolutional neural network model based on Vgg16 is used in the artificial intelligence algorithm...An artificial intelligence enhanced star identification algorithm is proposed for star trackers in lost-inspace mode.A convolutional neural network model based on Vgg16 is used in the artificial intelligence algorithm to classify star images.The training dataset is constructed to achieve the networks’optimal performance.Simulation results show that the proposed algorithm is highly robust to many kinds of noise,including position noise,magnitude noise,false stars,and the tracker’s angular velocity.With a deep convolutional neural network,the identification accuracy is maintained at 96%despite noise and interruptions,which is a significant improvement to traditional pyramid and grid algorithms.展开更多
基金Supported by the National Key Basic Research and Development (973) Program of China (No. G2000077606 )
文摘A star identification algorithm was developed for a charge-coupled device （CCD） or complementary metal-oxide-semiconductor （CMOS） autonomous star tracker to acquire 3-axis attitude information for a lost-in-space spacecraft. The algorithm took advantage of an efficient on-board database and an original “4- star matching” pattern recognition strategy to achieve fast and reliable star identification. The on-board database was composed of a brightness independent guide star catalog （mission catalog） and a K-vector star pair catalog. The star pattern recognition method involved direct location of star pair candidates and a sim- ple array matching procedure. Tests of the algorithm with a CMOS active pixel sensor （APS） star tracker result in a 99.9% success rate for star identification for lost-in-space 3-axis attitude acquisition when the angular measurement accuracy of the star tracker is at least 0.01°. The brightness independent algorithm requires relatively higher measurement accuracy of the star apparent positions that can be easily achieved by CCD or CMOS sensors along with subpixel centroiding techniques.
基金the National Natural Science Foundation of China(No.6152403)。
文摘An artificial intelligence enhanced star identification algorithm is proposed for star trackers in lost-inspace mode.A convolutional neural network model based on Vgg16 is used in the artificial intelligence algorithm to classify star images.The training dataset is constructed to achieve the networks’optimal performance.Simulation results show that the proposed algorithm is highly robust to many kinds of noise,including position noise,magnitude noise,false stars,and the tracker’s angular velocity.With a deep convolutional neural network,the identification accuracy is maintained at 96%despite noise and interruptions,which is a significant improvement to traditional pyramid and grid algorithms.
