The most accurate star centroiding method for star sensors is the Gaussian fitting (GF) algorithm, because the intensity distribution of a star spot conforms to the Gaussian function, but the computational complexity of GF is too high for real-time applications. In this paper, we develop the fast Gaussian fitting method (FGF), which approximates the solution of the GF in a closed-form, thus significantly speeding up the GF algorithm. Based on the fast Gaussian fitting method, a novel star centroiding algorithm is proposed, which sequentially performs the FGF twice to calculate the star centroid: the first FGF step roughly calculates the Gaussian parameters of a star spot and the noise intensity of each pixel; subsequently the second FGF accurately calculates the star centroid utilizing the noise intensity provided in the first step. In this way, the proposed algorithm achieves both high accuracy and high efficiency. Both simulated star images and star sensor images are used to verify the performance of the algorithm. Experimental results show that the accuracy of the proposed algorithm is almost the same as the GF algorithm, higher than most existing centroiding algorithms, meanwhile, the proposed algorithm is about 15 times faster than the GF algorithm, making it suitable for real-time applications.

In this article, we consider the problem of extracting star spots for traditional star sensors under highly dynamic conditions. Under this setting, it is difficult to effectively extract star spots with a low signal-to-noise ratio (SNR), resulting in a failure to estimate the attitude of star sensors. We propose a method named optimal directional connected component (ODCC) for this task. According to the dynamic characteristics of star sensors, we propose an image enhancement method that can adaptively estimate the directions of star spots and integrate the star image so that the SNR of the star spots is increased. This aids in searching for spots with a low SNR. According to the imaging characteristics of a star, we model the imaging region of a star as a rectangle and then creatively cast the problem of extracting spots into a problem of estimating the endpoints of the rectangle. A maximum likelihood estimation method based on the Gaussian mixture model is proposed to determine the region of the star spot. Experiments demonstrate the extraction capability and accuracy of the method. In conclusion, this method has a promising application value for improving the dynamic performance of traditional star sensors.

A fast and robust star identification algorithm is proposed in this paper. The algorithm is derived from the grid algorithm but significantly improves the robustness by increasing the probability of the correct close neighbor star and reducing the pattern noise. Firstly, star pair patterns are constructed rather than star patterns. And a traversal method is adopted to find all the possible closest neighbor stars, which improves the robustness towards brightness noise. Secondly, the two-dimensional angular distance features are used instead of grid features in Cartesian coordinates to improve the robustness towards positional noise. The simulation results show that the proposed algorithm is quite robust to a variety of noise conditions. The identification rate of the proposed algorithm is higher than the grid algorithm and similar to the pyramid algorithm, while the identification speed is up to dozens times faster than the pyramid algorithm.

In this paper, a novel guide star catalog generation algorithm is presented, which is a crucial part of an advanced star tracker design since the performance and reliability of star identification and attitude determination depend on the guide star selection. First, we propose an analytical method to estimate the stellar instrument magnitude and characterize the associated errors. Those stars whose instrument magnitudes are higher than the star tracker sensitivity constitute the original catalog. Then, the probability model of each star in the original catalog for attitude determination is established as the function of the field of view (FOV), the brightness accuracy, and the mean number of stars in FOV based on the brightness and uniformity principles, and the stars with larger probability are selected as guide stars. The accuracy of photometry method is validated by estimating the Hipparcos catalog instrument magnitude and comparing them with actual data. The efficiency of guide star selection algorithm is demonstrated by generating guide star catalog for a star tracker with a 16° × 16° FOV.

Star identification is a key technology in the research of star sensors. As a classical star pattern for identification, the radial pattern of a chosen star represents the geometric distribution of neighboring stars. Previous researches usually combine the rotation-invariant radial pattern with other less reliable patterns like cyclic pattern to identify the chosen star. Seldom studies use the radial pattern for direct identification as it is sensitive to noise and inefficient to eliminate spurious matches. To solve the problem, a novel star identification algorithm based on the recommended radial pattern was developed. The algorithm introduced the Apriori algorithm to mine the association rules between a star and neighboring stars, recommend the companion stars, and generate the recommended radial pattern. Based on the association rules, the algorithm integrated the matching results of a star and recommended companion stars to identify the chosen star. The simulation test and night sky image test both show that the proposed algorithm is robust to position noise, brightness noise, and false stars. The novel matching strategy is not limited to the radial pattern algorithms, it provides a new idea for all the star pattern algorithms.

Dynamic performance is an important index of star sensors. When the star sensor is under dynamic conditions, imaged star spots stretch to steaks as the energy is more dispersive than that of static conditions, which leads to decreases in the signal-to-noise ratio (SNR) of the star spots. The threshold method is unable to detect enough stars under dynamic conditions. However, star sensors need sufficient stars to guarantee normal work. The image enhancement method enhances the star spots by integrating along with their movement directions. But it still misses a lot of low-SNR stars as the determination of movement directions is sensitive to noise, which leads to the limited enhancement of low-SNR stars. To solve the problem, this study presented an improved UNet-based star spot extraction method for star sensors under dynamic conditions. The improved UNet, defined as SSA-UNet, used a star streak attention module (SSAM) to help it focus on the star spots and improve the extraction performance. The presented method used SSA-UNet for the rough extraction and modified the rough extraction in the subsequent process. Both simulation and ground experiments proved that the proposed method can accurately extract more low-SNR star spots than the image enhancement method, which had great significance for further improving the dynamic performance of star sensors.