The Xray pulsar can provide navigation information of the position, velocity, attitude and time with highprecision autonomously for the spacecraft in the deepspace and nearEarth orbit. But for the spacecraft around the Earth, the pulsars are sometimes invisible due to the eclipse of the central body, so the visibility of pulsars becomes a key factor on the nearEarth navigation application of the Xray pulsars navigation (XPNAV). This paper analyzes the visibility of 30 pulsars picked for the 7 orbits of the Beidou System. The simulation results of the number of the visible pulsars in one day are obtained and the situations of the eclipse of the Sun, Erath and Moon in one year are also presented. The rules of the influence of the celestial bodies on the visibility of pulsars are summarized in this paper, which can provide a data support for the autonomous orbit determination and time keeping of the Beidou System by using the XPNAV.

Target detection and tracking technology have been widely used in the fields of transportation, medical, safety and military affairs, etc. However, there still exist some challenges in target detection and tracking, such as dim small target, complex background, target occlusion, and appearance changes, etc. On the other hand, as the most effective biointelligence system, Human Visual System has significant advantages in image processing. In this paper, combining neural engineeringoriented brainlike models and computer engineeringoriented DNNs, three target detection and tracking algorithms based on braininspired models and DNNs are proposed, including: a moving target detection algorithm based on Algorithmic Lateral Inhibition (ALI) model, a dim target detection algorithm based on StructureContrast Visual Attention model, and a target tracking algorithm based on memory mechanism and convolutional feature. The comparison experiments show that applying the braininspired models and DNNs to the infrared target detection and tracking is beneficial to achieve accurate target detection and robust tracking under complex conditions.

The rotation parameters of a small body are scientific data, as well as basic data for small body mapping and landing navigation. A small body rotation parameter estimation method is studied based on image feature points tracking and extended Kalman filter in the process of small body exploration approach. The method first establishes a small body's rotation relationship model, which represents the attitude change of the small body in the camera coordinate system; then defines the state equation of the small body's rotation, and pushes to the calculation process of the extended Kalman filter. By tracking the feature points on the observation image sequence, the extended Kalman filter method is used to obtain the estimation of the rotation axis and the rotation angular velocity of the small body. In the experiments, when the distance between the simulated camera and the small body camera is 100km, the initial value of the quaternion between the camera coordinate system and the small body coordinate system, the tracking accuracy of the feature points on the image, and the camera's observation direction are analyzed for the rotation parameters of the small body. The experimental results show that the proposed method can obtain highprecision estimation results.

Infrared (IR) saliency detection with high detection accuracy is a challenging task due to the complex background and low contrast of IR images. In this paper, an IR saliency detection method based on a new visual attention framework is proposed, which comprises two phases. In the first phase, a Gray & Contrast Features (GCF) model is established, in which the IR image is processed in two feature channels, a gray feature channel and a contrast feature channel. And then a primary feature map can be obtained by fusing the gray and contrast features from these two channels, which is the basis of the second phase. In the second phase, a Similaritybased Bayes (SB) model is established, in which two prior probabilities and two likelihood functions are calculated according to the previously obtained primary feature map. Finally, the saliency map is calculated with the obtained prior probabilities and likelihood functions by Bayes formula. Experimental results indicate that the proposed method can effectively reduce noise and enhance contrast of IR images with complex background and low contrast, and obtain a higher detection accuracy and robustness than seven comparison algorithm methods.

Femtosecond laser direct writing technology has significant advantages in the field of complex threedimensional microstructure processing, and the accuracy of the focusing directly affects the integrity of the processed structure.This paper presents an image focusing technique in which light source and object are temporarily placed in the optical path. By adjusting the position of the object, the imaging plane is consistent with the laser focusing plane, so that the laser focusing state can be indirectly reflected by the clear and resolvable imaging state. The Zemax software is used to simulate the processing optical path and the focusing optical path respectively. The results show that the two kinds of simulation results can achieve the same back working distance and good imaging quality, which proves the feasibility of the method. The analysis shows that the imaging error of this process is mainly caused by the focal depth of the imaging lens (3.9 μm), and the ideal focusing accuracy can reach within 1/2 focal depth. A twolayer cylindrical structure with a singlelayer height of 5μm is designed for testing. It is verified through multiple experiments that the height error of the processed components is within the range 1.5 μm, which is consistent with the theoretical analysis and meets the focus requirements of the femtosecond laser system.

To improve the safety of lunar rover during autonomous detection, an improved path planning algorithm based on Probabilistic Roadmap (PRM) is proposed. The algorithm is based on the distance transform map to improve the sampling method of the PRM algorithm. By controlling the sampling point away from obstacles, the planned path is kept away from obstacles, avoiding dangerous situations of advancing close to obstacles, and improving the safety of lunar rover. To evaluate the safety of the path, two safety indicators, a safety alert factor and a minimum safety alert factor, are proposed. And the paths generated by A*, PRM and the improved PRM algorithm are evaluated for safety in the lunar surface simulation environment. The results show that the improved PRM, compared with A* algorithm, the safety alert coefficient and the minimum safety alert coefficient of the algorithm are increased by 2.20m and 1.00m respectively; compared to the PRM algorithm, the safety alert coefficient and the minimum safety alert coefficient are increased by 1.68m and 1.00m, respectively. The improved PRM algorithm is not only limited to the path planning of lunar rover, but can also be applied to exploration robots and selfdriving cars that require high path safety.

Estimating the relative poses of satellites to spacecrafts is a supporting technology in many space manipulation activities. Addressing the single light source in the space and the lightreflecting surfaces of the satellites, a novel binocular method is proposed for satellite pose estimation in this paper. When the panel landmarks of the satellites can be observed, the proposed method uses the outer docking rings and the panel landmarks for pose estimation, and otherwise, the ORB features are used to measure the satellite poses. In addition, the opticalflow tracker and the Kalman filter are used to optimize the estimation results for improving the robustness of the proposed method. Simulation results demonstrate that this method can calculate the relative poses of the satellites accurately in the singlelightsource space.

Space rendezvous and docking technology is a prerequisite for onorbit services such as assembly, recovery, resupply, maintenance, and crew exchange or rescue of space shuttles, space platforms, and space transportation systems. Cameratype rendezvous and docking sensor (CRDS) is the only sensor which can indicate sixdegreeoffreedom (6DOF) information about relative position and attitude between two vehicles over the final approach phase. And it is comprised of the camera mounted on the tracking aircraft and cooperative target clusters on the target aircraft. The relative position and attitude information can be obtained, adopting the corner reflectors as cooperative targets. Requirements analysis, design analysis, and the results of sensor performance testing with the targets are discussed in detail. The design was successfully applied in autonomous rendezvous and docking missions of SZ11&TG2 and TZ1&TG2. In the future, the design will be revised to better fit the missions such as KJZ and GXC.

Dust particles are generated with the satelliterocket separation by explosion of initiating devices. Star trackers are interfered by scattered stray light from the particles, resulting in attitude capture failures. In this paper, imaging with scattered stray light is modeled and analyzed, and the common feature of star energy distribution between star images with and without stray light is proposed. A fast star extraction algorithm based on crosscorrelation coefficient is presented, which is effective in conditions of strong scattered stray light. Based on the algorithm, the star points are extracted practically, and the attitude capture is realized rapidly. Simulations on star extraction and attitude capture with different stray light conditions are carried out. In comparison with the traditional algorithm, the proposed algorithm shows advantages on star extraction capability and attitude capture in different conditions.

Due to the inherent defects of laser sensors, the original point cloud of noncooperate targets is usually irregularly distributed, which brings great challenges to highquality 3D surface reconstruction of noncooperate targets. In this paper, we propose a local hierarchical clustering method based on global constraints to improve the consistency of noncooperative target point cloud distribution. Specifically, our method includes two main steps. The first one is the adaptive octreebased 3D spatial decomposition with global constraints and the second one is hierarchical clustering based on global constraints. The main purpose of the former one is to reduce the complexity of the algorithm, and the later one aims to convert the nonuniform point set to uniform one. Experiments are carried out on three noncooperative target models. The results of visualization and quantitative calculation verify the effectiveness of our method.

An infraredvisible binocular camera can not only perceive the temperature of the targets, but also obtain the color and texture information of them, which has gotten more and more application in daily life and industrial production. Considering the different imaging mechanism between infrared and visible camera, different calibration targets are needed in general calibration. In this paper, we propose a pixellevel fast calibration method for the infraredvisible binocular system. By designing a unique circular target rig, onestep calibration is realized. The calibration precision is below 1 pixel in visible camera pixel coordinate system. This method can be used to the large batch and low cost civil infrared production calibration system.