Three-dimensional (3D) surface models, e.g., digital elevation models (DEMs), are important for planetary exploration missions and scientific research. Current DEMs of the Martian surface are mainly generated by laser altimetry or photogrammetry, which have respective limitations. Laser altimetry cannot produce high-resolution DEMs; photogrammetry requires stereo images, but high-resolution stereo images of Mars are rare. An alternative is the convolutional neural network (CNN) technique, which implicitly learns features by assigning corresponding inputs and outputs. In recent years, CNNs have exhibited promising performance in the 3D reconstruction of close-range scenes. In this paper, we present a CNN-based algorithm that is capable of generating DEMs from single images; the DEMs have the same resolutions as the input images. An existing low-resolution DEM is used to provide global information. Synthetic and real data, including context camera (CTX) images and DEMs from stereo High-Resolution Imaging Science Experiment (HiRISE) images, are used as training data. The performance of the proposed method is evaluated using single CTX images of representative landforms on Mars, and the generated DEMs are compared with those obtained from stereo HiRISE images. The experimental results show promising performance of the proposed method. The topographic details are well reconstructed, and the geometric accuracies achieve root-mean-square error (RMSE) values ranging from 2.1 m to 12.2 m (approximately 0.5 to 2 pixels in the image space). The experimental results show that the proposed CNN-based method has great potential for 3D surface reconstruction in planetary applications.

Rocks on the Martian surface exhibit signatures of geological activities. Rock distribution is also an important factor for landing site selection for exploration missions. This study investigated the rock abundance and erosion rate at the landing site of the Chinese Mars rover, Zhurong, in southern Utopia Planitia. A convolutional neural network based method was developed for rock detection from the images (0.7 m/pixel) obtained by using the high‐resolution imaging camera (HiRIC) onboard the Tianwen‐1 orbiter. Approximately 10,000 rocks with a minimum diameter of 1.4 m (2 pixels on the image) were extracted from the HiRIC images covering the Zhurong landing site and its neighboring region. According to the results, the rock abundance at the Zhurong landing site was approximately 5%. Measurements from the ground images acquired by the Zhurong rover after landing confirmed a rock abundance of approximately 5% at the landing site, consistent with the orbital measurement before landing. Furthermore, we investigated the rock erosion rate at the Zhurong landing region. Rock abundance was negatively correlated with the crater retention age and positively correlated with crater diameter. The obtained rock breakdown rate (0.054 nm/y to 0.074 nm/y) and surface denudation rate by filled volume of craters (0.104 nm/y to 0.209 nm/y) were in agreement with the values previously reported for the Amazonian period on Mars. The low erosion rate implies a dry climate and the absence of fluvial erosion processes in the Amazonian period.

This study presents a global database of pitted cones on Mars. This database was created using an active deep-learning approach based on a global mosaic of Mars Context Camera (CTX) images, followed by extensive manual verification. The global database contains 250,547 pitted cones along with their central coordinates and diameters. 94.2% of the pitted cones have diameters ranging from 100 m to 1 km. Most of them are in the northern lowlands, while some are scattered in the southern highlands. Systematic evaluation of the extracted pitted cones is performed in Isidis Planitia, Acidalia Planitia, and Cerberus Palus on Mars. The results reveal that the precision ranges from 79.2% to 85.6%, and the recall rate ranges from 62.4% to 87.7%, indicating a varying among regions but satisfactory performance. Manual digitizing evaluation indicates that over 94.1% of manual labels by different annotators have a differentiation rate of 10%, showing promising precision and consistency. This global database can facilitate further research on the origin and evolution of pitted cones on Mars.

This article tackles the visual localization of unmanned aerial vehicles (UAVs) in the presence of multisource and cross-view images are involved. We present a lightweight end-to-end scene graph encoding and matching network that finds the best matches for the airborne camera views from the reference image maps. The scene graph addresses the challenges of encoding the semantic scene by aggregating the image convolutional features into global and structured semiglobal descriptors. The principal contributions of this article are as follows: First, we develop a new network architecture that embeds a nonlocal block and a modified vector of locally aggregated descriptors network (NetVLAD) into a backbone convolutional neural network. The main component of the modified NetVLAD is a cluster similarity masking graph (CSMG) encoder, which is proposed to replace the feature-cluster residuals computing in NetVLAD with cluster consensus feature aggregation and structure-aware scene graph extraction. In addition, a global descriptor is extracted by a nonlocal block to label each image with a discriminative global feature descriptor. Second, we develop a new triplet loss for the network training procedure to learn the features at different semantic levels. The proposed global descriptor and CSMG encoder are trained together according to a weighted sum of cosine triplet losses. Third, the global descriptor from the nonlocal block and semiglobal descriptor from the CSMG encoder work hierarchically for coarse-to-fine image retrieval and can achieve real-time efficiency and favorable accuracy of image searching and matching from the reference image map. We train and test the model on two challenging benchmark datasets. We also test the pretrained model on a dataset collected by a fixed-wing UAV to further evaluate the model's generalizability. The benchmark evaluations and ablation experiments show that the developed method outperforms state-of-the-art methods and achieves superior performance in the real-time matching of UAV images and reference image maps for UAV visual localization. Open-source code is available on GitHub.