<p><strong>Abstract.</strong> Point cloud classification is quite a challenging task due to the existence of noises, occlusion and various object types and sizes. Currently, the commonly used statistics-based features cannot accurately characterize the geometric information of a point cloud. This limitation often leads to feature confusion and classification mistakes (e.g., points of building corners and vegetation always share similar statistical features in a local neighbourhood, such as curvature, sphericity, etc). This study aims at solving this problem by leveraging the advantage of both the supervoxel segmentation and multi-scale features. For each point, its multi-scale features within different radii are extracted. Simultaneously, the point cloud is partitioned into simple supervoxel segments. After that, the class probability of each point is predicted by the proposed SegMSF approach that combines multi-scale features with the supervoxel segmentation results. At the end, the effect of data noises is supressed by using a global optimization that encourages spatial consistency of class labels. The proposed method is tested on both airborne laser scanning (ALS) and mobile laser scanning (MLS) point clouds. The experimental results demonstrate that the proposed method performs well in terms of classifying objects of different scales and is robust to noise.</p>

Abstract. Navigation from LEO satellites own many merits and attracts increasing popularity recently. In addition to increasing the signal availability, the low signal strength loss and fast satellite geometry change from LEO satellite are particularly appealing in challenging environments. Recently, a few researchers attempt to navigate with non-cooperative signals from LEO satellites with pure phase lock loop (PLL) or frequency lock loop (FLL), while a more practical solution to utilizing LEO navigation is joint positioning with the existing GNSS signals which has not been seriously studied. In this study, we proposed a joint GPS and LEO navigation signal tracking strategy that employs a vector tracking loop (VTL) with fully considering the high dynamic characteristics of the LEO signals. In order to solve the high dynamics problem, the second-order deviation parameters were considered in the extended Kalman filter, which is more adaptive to the non-linear variation of the signal acceleration. In addition, a carrier-to-noise ratio (C/N0) based observation noise determination strategy is employed to adapt different observation conditions. The proposed method was verified with different simulation data and the results indicate the adaptive vector tracking loop is capable of tracking GPS and LEO signals simultaneously and robustly. The benefit is particularly in the weak signal scenarios. The experiment results also reveal that the joint vector tracking loop improves positioning accuracy in GNSS challenging environments.

Abstract. Non-spherical gravity plays a crucial role in the LEO satellite orbit determination and prediction. In recent years, several new gravity models have been proposed with more comprehensive ground and space-borne data. The impact of the gravity models has been extensively studied while its impact on the orbit prediction has not attracted enough attention. With the risen of the mega LEO constellation, new applications such as the LEO navigation requires real-time precise orbit, which increases the importance of the precise orbit prediction. In this study, we selected six popular gravity models, namely JGM3, EGM2008, EGM96, EIGEN2, GL04C, and GGM03S, and compared their performance in different LEO orbit predictions. The comparison results indicate that there is no single optimal gravity model for all LEO orbit prediction scenarios. For short-term prediction, JGM3, EGM2008, GL04C models perform better while in long-term prediction JGM3, EGM96, EIGEN2 have more potential. The results also reveal that the optimal model changed with time. In addition, the impact of the gravity order on the orbit prediction is investigated, the results indicate that for satellites with lower orbital heights, the gravitational field order required to achieve a certain truncation error is higher than for satellites with higher orbital heights. The authors also explore the effect of gravitational field-associated permanent tides on orbital prediction. In one day, for satellites with an orbital altitude of about 970km, the effect of permanent tides on 3D RMS is 6.92m; for satellites around 710km, the effect of permanent tides on 3D RMS is 4.20m; for satellites around 970km, the effect of permanent tides on 3D RMS is 2.07m.