3 October 2022 in IEEE Sensors Journal
In order to meet the urgent needs of short-wave infrared (SWIR) star sensors in near-ground navigation applications, starting from the detection capability and detection accuracy, a design method for the optical system of near-earth space star sensor is proposed, including the selection method of optical parameters and the determination method of structure parameters. Based on the presented method, the index requirements are given, and an H-band catadioptric star sensor optical system with a focal length of 1174mm, an aperture of 200mm, and an entire field of view (FOV) of 0.8° is designed. The relative distortion is lower than 0.02%, the chromatic aberration is less than 0.1 pixels, over 90% of the star point energy is distributed in 3 × 3 pixels, the ambient temperature adaptability is acceptable, the expected consequence is achieved, and the feasibility of the method is proved. Furthermore, the stray light elimination system is designed to avoid the influence of direct sunlight on stellar imaging, and an integrated navigation system is designed to reduce interference of solid sky background on stellar detection. The derived information of sky background radiation, degree of polarization and angle of polarization, is used to improve navigation reliability. This paper can provide a theoretical foundation and technical support for the engineering application of star sensors in tanks, ships, aircraft, and missiles.
14 September 2021 in Neural Computing and Applications
2 November 2022 in Infrared Physics & Technology
10 September 2021 in arXiv
To eliminate the problems of large dimensional differences, big semantic gap, and mutual interference caused by hybrid features, in this paper, we propose a novel Multi-Features Guidance Network for partial-to-partial point cloud registration(MFG). The proposed network mainly includes four parts: keypoints' feature extraction, correspondences searching, correspondences credibility computation, and SVD, among which correspondences searching and correspondence credibility computation are the cores of the network. Unlike the previous work, we utilize the shape features and the spatial coordinates to guide correspondences search independently and fusing the matching results to obtain the final matching matrix. In the correspondences credibility computation module, based on the conflicted relationship between the features matching matrix and the coordinates matching matrix, we score the reliability for each correspondence, which can reduce the impact of mismatched or non-matched points. Experimental results show that our network outperforms the current state-of-the-art while maintaining computational efficiency.