The Grove Mountains are located in the hinterland of the Princess Elizabeth Land of the East Antarctic Ice Sheet. 3D modelling of the Grove Mountains is highly important for Chinese scientific research in the Antarctic. Through this research, a 3D model of the entire 3200 km2 Grove Mountains area and elaborate 3D models of the 64 nunataks were established. A 3D modelling method for the Grove Mountains based on multisource high-resolution stereo satellite images is proposed. This method includes four main steps: specific strategies for snow-covered and blue ice textures, a heterogeneous image fusion process, multi-image combined adjustment and patch-based multiview stereo (PMVS). The results of the experiments reveal that the bias, mean absolute error (MAE) and root mean square error (RMSE) were the smallest for the proposed DEM and the (Reference Elevation Model of Antarctica) REMA. These findings indicates that the proposed DEM is most similar to the REMA. The proposed DEM and the REMA had the strongest correlation. The correlation coefficient r was 0.99906. The absolute accuracy of the proposed DEM was verified by using the actual measured GPS points. The minimum elevation error was −0.4972 m, which occurred in the North Gale Escarpment.

Asteroid 3D modelling is of great significance for asteroid exploration mission. Asteroid 3D modelling method based on the combined adjustment of stereo images and laser altimetry data is proposed. The methodological novelty is: (1) 3D modeling pipeline for asteroids is proposed, include: laser footprint positioning, laser point image coordinate extraction, combined adjustment and PMVS (Patch based Multi View Stereo). (2) combined adjustment model integrating stereo image and laser altimeter data has been established. The boresight misalignment and the eccentric distance of the main camera, the laser altimeter pointing angle, are added as additional parameters for improved accuracy. The accuracy and effectiveness are verified by simulated experiment. Through verification, the highest RMS (Root Mean Square) of laser point positioning is 0.85 mm, and the lowest RMS is 35.18 mm. The checkpoints RMS for 3D modelling, the highest is 4.55 mm and the lowest is 11.66 mm. Through the linear fitting analysis, it is determined that the modeling accuracy is 0.16 m at a photography distance of 500 m and 0.31 m at a photography distance of 1000 m. Through calculation, the Pearson correlation coefficient reaches 0.97. This shows that the proposed model has a high correlation with the original model.

Wild camels living in northwestern China and Gobi A of Mongolia are endangered. This research indicates that the amount of suitable habitat for wild camels will continue to shrink and become isolated in the future. These results are based on our analysis of continuous independent satellite tracking data obtained for 37 peak wild camels in China and Mongolia from 2012 to 2025. Based on the MaxEnt model, a suitable habitat model is established, and future trends are predicted. Currently, the area of suitable habitat totals 96,116.69 km2. Under the newest CMIP6 (Coupled Model Intercomparison Project Phase 6) climate model, the predicted suitable habitat area will decline to 3159.44 km2 by 2030, 169.07 km2 by 2050 and 0 km2 by 2070. The accuracy of the established model is verified by using the AUC, the TSS, the kappa coefficient and a cross-validation experiment. Owing to the influences of climate change and anthropogenic activities, wild camels will face severe survival challenges in the future. We suggest the establishment of the Sino-Mongolia Wild Camel United Nature Reserve. The overall level of protection for wild camels should be strengthened, and the degrees of movement and population exchange exhibited by wild camels in the north-south direction should be increased.

野骆驼(Camelus ferus),又名野双峰驼,是世界上唯存的骆驼科(Camelidae)真驼属野生种,属于我国国家一级重点保护野生动物,被IUCN红色名录列为极度濒危物种.野骆驼生性机警,栖息在自然环境极端恶劣的荒漠半荒漠地区,耐严寒酷暑,耐饥渴,不畏风沙,极端干旱环境下的适应性和抗逆性强,能饮用高矿化度苦咸水,是其所在荒漠生态系统的旗舰物种.

Self-calibration method for the stereo vision system of the Chang'E-5 probe is proposed. The footprints left by the robot arm after touching the ground are regarded as the lunar control points. A stereo constraint is proposed, and a combined adjustment model is established. The accuracy of the proposed method is verified by Earth-based and lunar experiments. In the Earth-based experiment, the proposed method obtained a 3.18 mm average positioning error in 50 sets of tests, the average residual error of nine footprints was 0.45 pixels, and the unit weight mean square error of Moncam B was 0.181 pixels. In the lunar experiment, the proposed method obtained a 3.22 mm average positioning error in 50 sets of tests, and the unit weight mean-square error of Moncam B was 0.172 pixels. The experimental results show that the proposed method provides high accuracy and stability.

Satellite tracking and positioning technology has been widely used to estimate species home ranges, habitat selection, and migration behavior, which can provide the scientific basis necessary for endangered species conservation and management. In this study, the home ranges of two wild camels（Camelus ferus） were determined using minimum convex polygons, Kernel density estimators, and dynamic Brownian Bridge Movement Models（dBBMM）. The advantages,disadvantages, and applicable scenarios of these methods were discussed and key habitats and conservation priority areas of wild camels were identified. The dBBMM home range was the most accurate in relation to the distribution of the camels studied. The average home range（95% dBBMM） and the average core home range（50% dBBMM） were 300. 11 km2and 7. 02 km2, respectively. The average monthly home range（95% dBBMM） of the two wild camels was 164. 98 km2and 39. 67 km2, respectively, and the average monthly core home range（50% dBBMM） was 4. 69 km2and 2. 71 km2, respectively. The northern and western regions of the nature reserve harbored key habitats for wild camels, and should be regarded as conservation priority areas. We propose the use of a multi-method estimation to comprehensively and holistically identify key habitats. The dBBMM home range calculations of wild camels at different life history stages can identify their stopover sites and corridors, which is helpful for the development of management measures and fine-scale conservation actions.

野骆驼(Camelus ferus)天性警惕,人们很难见到,研究监测难度相对较大.常用的监测方法主要有红外线相机和卫星定位跟踪项圈."无线网桥+雷达摄像装置"首次用于野骆驼监测.无线网桥监控能够减少因监测人员在监测区域内频繁活动给野生动物带来的干扰,较完整地监测到野生动物在 自然状态下的活动状况,且能够全天候不间断监测,数据保存完整,监测区域覆盖面大、获取物种种类多、影像质量高等.2020年9月至2021年8月期间,安南坝保护区在此监测区域共监测到国家重点保护野生动物18种,有野骆驼(Camelus ferus)、雪豹(Cuon alpinus)、豺(Panthera uncia)、高山兀鹫(Gyps himalayensis)、黑颈鹤(Grus nigricollis)等国家Ⅰ级保护野生动物8种,其中在库姆塔格沙漠南缘山区首次发现雪豹.

"祝融号"是我国首个成功在火星表面开展巡视探测的火星车.为了确保火星车遥操作任务精准的实施,需要对立体视觉系统实施自检校.本文提出了附有立体约束和桅杆机构运动约束的联合平差自检校模型.经过模拟试验证明本文提出的自检校方法具有较高的视觉定位精度.在火星真实试验中,获得检核点中误差的平均值为12.3 mm.有效保障了"祝融号"火星车精准地完成遥操作任务.

野骆驼,又名野双峰驼,偶蹄目,骆驼科,真驼属动物,在中国属于国家一级重点保护野生动物,在《世界自然保护联盟濒危物种红色名录》中被列为极度濒危物种(CR),有陆地脊椎动物活化石之称.

Tianwen-1 probe is equipped with the descent camera. Descent images are very special and characteristic. Many image processing missions can be done when parameters of descent camera are known. Therefore, the self-calibration method for descent camera based on the structure from motion (SfM) is proposed. The relative orientation model of the oblique and the vertical baselines is proposed in order to provide the accurate initial value. Simulated experiment of the unmanned aerial vehicle in the desert environment verifies that the proposed method has high positioning accuracy and the stability. Finally, the proposed method is practical used for Tianwen-1 mission. The average value of the root-mean-square error of the in-orbit checking points is 0.875 m.

The parachute of the probe plays an important role in the deceleration of the Mars landing mission. This paper provides an overview of the visual measurement system of the parachute in Tianwen-1 exploration mission. Using two monitoring cameras, the system has captured stereo images during the parachute deployment. The morphological and motion parameters of the parachute are reconstructed by image-based feature tracking and 3D reconstruction methods. To improve the image quality, a super-resolution (SR) reconstruction method based on sparse dictionary coding is proposed. In addition, we use a multi-scale feature tracking method to track the feature points and update the motion parameters. The system has been verified by the test on the Earth and successfully applied on the onboard images. The reconstructed parachute parameters are important in the post-flight performance assessment, and the research results can be used as a regression analysis means for the structural design.

There is a pair of binocular cameras for navigation onboard the Chang’e-4 lunar rover. Although the cameras were calibrated before launch, it is necessary to calibrate them again after landing on the moon, especially the external parameters. In this article, an image of a solar panel containing parallel-line features is used for recalibration. According to the collinear equations of an image point, object point and projection centre, the algebraic relationship between the slope k and intercept t parameters of a line in the image and the external parameters of the cameras are deduced. Through this algebraic relationship, we propose an external parameter recalibration method based on an image that includes parallel lines on a plane in the object space. Three experiments were carried out to evaluate the effectiveness and reliability of the proposed method.

Wheeled ground robots are widely used in planetary exploration missions, where slip can measure problems during driving and pre-acquisition slips play an important role. The current best traditional model is a predictive model based on visual sliding learning, which is divided into two steps: determining the terrain type and constructing a geometry-based sliding model. Since the geometry-based sliding mode model is not realistic in the orbit learning, this paper introduces the possibility of applying geometric models to planetary missions. In order to conduct a more comprehensive study of geometry-based models, three new related variables have been introduced. Then the geometric model-based rain map rover prototype data obtained from the indoor experiment is applied to the rain map rover data in the Lunar Changchun No. 3 mission. The results show that terrain geometry is the main influencing factor when the simulation of local shaped materials reaches a certain level. This means that the geometry-based slip prediction model has the potential to be applied under similar conditions, and the results of the article have certain value. At the same time, this method provides another way of thinking for China’s future Mars project.

Currently, deep neural networks have gained impressive performance for object detection when plenty of labeled training data are available. However, a substantial quantity of labeled data is typically rare and time-consuming to obtain in deep space exploration. To possibly alleviate this problem, we use domain adaptation (DA) to effectively detect unannotated real data samples in the crater detection problem with the help of autoannotated synthetic data samples only. Specifically, we present a novel network, namely, CraterDANet, which unifies image- and feature-level adversarial DAs to bridge the domain gap between synthetic and real data. To evaluate our method, we present a new lunar crater dataset that contains nearly 20 000 small-scale craters captured from different Lunar Reconnaissance Orbital (LRO) Narrow Angle Camera (NAC) grayscale images that have large variations in shape, size, overlap, degradation, and illumination conditions. To the best of our knowledge, this is the first lunar crater dataset with diversification of illumination conditions, which provides bounding box annotations. Through experiments, the proposed CraterDANet, which was trained on labeled synthetic data without using any additional real labeled data, achieved an $F1$ -score of 80.27%, which is slightly better than traditional supervised methods trained on real data. Although the performance still falls short of the state-of-the-art Faster R-convolutional neural network (CNN), it achieves an $F1$ -score of +5.98% performance improvement compared with Faster R-CNN trained on synthetic data only. This result demonstrates the effectiveness of the proposed method for reducing the domain gap between synthetic and real data.

The hand-eye system is an important component of the Zhurong rover, which is crucial for positioning, terrain reconstruction, stereo vision measurement and mission planning functions. The hand-eye system consists of a navigation camera and mast mechanism, and its positioning accuracy has an important effect on the teleoperation mission of the Zhurong rover. A positioning method that combines close-range photogrammetry and robot kinematics is proposed. It includes geometric calibration, stereo vision measurement and coordinate transformations. Experiments were carried out in different scenes (for example, field, laboratory and Mars) to verify the positioning accuracy of the hand-eye system. The minimum and maximum positioning errors of the hand-eye system, as verified by in-orbit checkpoints, were 16.5 and 28.7 mm, respectively. This positioning accuracy can effectively satisfy the requirements of the teleoperation mission of the Zhurong rover.

Tianwen-1 is the first Mars probe launched by China and the first mission in the world to successfully complete the three steps of exploration (orbiting, landing, and roving) at the one time. Based on the unverifiable descent images which cover the full range of the landing area, trajectory recovery and fine terrain reconstruction are important parts of the planetary exploration process. In this paper, a novel trajectory recovery and terrain reconstruction (TR-TR) algorithm employing descent images is proposed for the dual-restrained conditions: restraints of the flat terrain resulting in an unstable solution of the descent trajectory and of the parabolic descent trajectory causing low accuracy of terrain reconstruction, respectively. A landing simulation experiment on a landing field with Mars-like landform was carried out to test the robustness and feasibility of the algorithm. The experiment result showed that the horizontal error of the recovered trajectory didn’t exceed 0.397 m, and the elevation error of the reconstructed terrain was no more than 0.462 m. The algorithm successfully recovered the descent trajectory and generated high-resolution terrain products using in-orbit data of Tianwen-1, which provided effective support for the mission planning of the Zhurong rover. The analysis of the results indicated that the descent trajectory has parabolic properties. In addition, the reconstructed terrain contains abundant information and the vertical root mean square error (RMSE) of ground control points is smaller than 1.612 m. Terrain accuracy obtained by in-orbit data is lower than that obtained by field experiment. The work in this paper has made important contributions to the surveying and mapping of Tianwen-1 and has great application value.

At 7:18 on May 15. 2021. Beijing time, more than 9 months after the launch of China's first Mars exploration mission Tianwen-1, the lander successfully landed in the pre-selected landing area in the southern Utopia Plain of Mars. The overall terrain in this area is relatively flat, making it easier for the Tianwen-1 lander to land safely, but there are still scattered obstacles such as craters, sand dunes, and ridges. Therefore, the Tianwen-1 lander still needs to be accurately located, in order to plan the path of the rover, preventing the rover from overturning due to obstructive terrain. Besides. incorporating the follow-up data of the rover into the unified Martian geographic coordinate system also needs the accurate landing site. The location of the landing site can be estimated according to the larder's landing orbit data, however, due to the parachute landing method, the error of the estimated location is large, so the image obtained after the landing of the Tianwen-1 lander should be used to accurately locate the landing site. Mars is far away from the Earth. and data transmission channel is limited. After the Tianwen-1 landed. the relevant image data received until May 22, 2021 only had the following three types of data: Remote sensing images obtained by the orbiter, a descent image during hovering obstacle avoidance and slow descent phase, and the ring-shot stereo image pairs of the Mars rover's navigation camera on the landing platform after landing. In order to locate the landing site of the Mars rover as soon as possible, this article proposes a precise positioning method based on the above limited data. At first. using the digital orthophoto map (DOM) generated by the orbiter remote sensing image as the base map, extract the craters, sand dunes, craters and ridges in the pre-selected landing area, and calculate the topological relationship between them. In addition, the topological relationship between the crater, sand dune, crater and ridge around the landing site is also calculated based on the ring-shot stereo image pair of the navigation camera. The second step is to perform fuzzy matching between the topological relationship of the feature geomorphology calculated by the navigation camera's ring-shooting stereo pair and the topological relationship of the feature geomorphology extracted from the DOM of the orbiter remote sensing image to obtain several suspected landing sites, and then to determine one of them as the initial position of the landing site based on the descent image during the hovering obstacle avoidance and slow descent phase obtained near the landing site. Finally, through the multi-image space resection of the navigation camera images, the precise position of the landing site is calculated as (109.925 degrees E, 25.066 degrees N). This method makes full use of the limited data at the initial stage of the landing, and realizes the first-time positioning of the Tianwen-1 lander's landing site. It has strong timeliness and ensures the safe movement of the rover. On June 6, 2021. U.S. time, the HiRISE camera on the Mars reconnaissance orbiter (MRO) captured high-resolution remote sensing images of the Tianwen-1 lander and the Mars rover. The position of the Tianwen-1 lander on the image is the same as that in this article, which further verifies the accuracy of the positioning method.

The navigation cameras of rover have been calibrated on the earth before launch, but it's a long flight before the rover arrives at another planet to work, during which the external force may cause the cameras' parameters to change. So, it is necessary to recalibrate the navigation binocular cameras after landing. This paper introduces a calibration method based on the solar panel as the calibration plate, which extracts the grid lines of the solar panel by Hough transform, the idea of clustering and least square line fitting method, and then recalibrates the external parameters of navigation binocular cameras according to the lines' parameters. At last, two experiments are carried out. In Experiment 1, the grid line extraction experiment is carried out with Chang'e 4 image, and the extraction method in this paper is more accurate compared with the traditional Hough transform line extraction results; In Experiment 2, the external parameters of binocular camera are calibrated with simulated linear parameters. The results show that the solution accuracy is high, and the method is feasible and effective. However, this camera calibration method takes the camera internal parameters as known values, and solves the external parameters of the left and right cameras separately. In the next step, we will continue to study the method of simultaneous calibration of camera internal and external parameters, and the method of one-step calibration of binocular camera.

During the three month movement of China's Zhurong Mars rover, it is a very important task to obtain the precise position of the rover and the terrain around the stations, which provides basic information for subsequent path planning, mechanism movement and scientific exploration. At the visual localization step, a bundle adjustment model by using the stereo navigation camera (Navcam) images with overlapping areas between adjacent stations can achieve continuous relative pose for the rover. which can effectively reduce the impact of wheel slip and inertial navigation system errors on positioning accuracy. Firstly, the initial position and attitude information of the Zhurong Mars rover and the rotation angle of the mast can be contained from the telemetry data. Then, through the inverse coordinate transformation, the Navcam pose in the landing north-east-down (NED) system can be obtained, which is regarded as the initial value of the Navcam's external orientation element. Secondly. the speeded-up robust features (SURF) matching algorithm is used to match stereo images at adjacent stations, and some sparse connection points are obtained. Thirdly, combined with the Navcam's internal orientation elements and relative pose parameters, a bundle adjustment model is constructed. A least square solution is used in the proposed bundle adjustment model, and the optimized Navcam pose at the current station can be calculated. Then, through the coordinate transformation. the Zhurong Mars rover's position in the NED system is obtained. At the terrain reconstruction step, a visual closed block network adjustment model and the stereo matching technology are used to obtain a digital elevation model (DEM) and orthophoto map (DOM). Firstly, an epipolar line correction method is performed on the Navcam's stereo images at the current camera station, and tie points are obtained by SURF. Secondly, the proposed closed block network adjustment model is constructed. When a total least squares solution is used in the proposed model, the optimized relative pose of the sequence images is calculated. Taking one pair of stereo images as a bridge, the sequence camera pose is integrated into the NED System. Thirdly, dense matching and forward intersection are performed in the current sequence images to obtain dense point clouds in the NED system. After point cloud dcnoising, interpolation and back projection. the DEM and DOM products are obtained, and the terrain reconstruction mission at the current station is completed. According to the ground experiment of the Zhurong Mars rover, the visual localization precision is better than 2%, the terrain reconstruction accuracy is 5 mm@3 m. In the orbit mission_ taking the visual localization results of the Zhurong Mars rover at the end of time months as an example, the cumulative error of the visual localization results is 5.1%. Based on the Navcam images sequence on the lander, Mars DOM images and one image from the descent camera with a height of about 200 m, the proposed technology is used to obtain the precise landing position (109.925 degrees E, 25.066 degrees N) of the Zhurong Mars rover on May 22, 2021. The landing position result corrects the estimated landing position of the rover based on the orbital parameters. The deviation of the two measured positions is about 2.1 km. At the same time, the localization and terrain reconstruction results effectively support on-orbit departure, path planning, and coordinate measurement of scientific target points, etc.

AbstractTo set the sleeping mode for the Yutu-2 rover, a visual pose prediction algorithm including terrain reconstruction and pose estimation was first studied. The terrain reconstruction precision is affected by using only the stereo navigation camera (Navcam) images and the rotation angles of the mast. However, the hazard camera (Hazcam) pose is fixed, and an image network was constructed by linking all of the Navcam and Hazcam stereoimages. Then, the Navcam pose was refined based on a multiview block bundle adjustment. The experimental results show that the mean absolute errors of the check points in the proposed algorithm were 10.4 mm over the range of $\boldsymbol{L}$ from 2.0 to 6.1 m, and the proposed algorithm achieved good prediction results for the rover pose (the average differences of the values of the pitch angle and the roll angle were −0.19 degrees and 0.29 degrees, respectively). Under the support of the proposed algorithm, engineers have completed the remote setting of the sleeping mode for Yutu-2 successfully in the Chang’e-4 mission operations.