Jiujiang Zhang

Planetary photogrammetry remains an important technical means of producing high-precision planetary maps. High-quality control networks are fundamental to successful bundle adjustment. However, current software tools used by the planetary mapping community to construct and edit control networks exhibit very low efficiency. Moreover, redundant and invalid control points in the control network can further increase the time required for the bundle adjustment process. Due to a lack of targeted algorithm optimization, existing software tools and methods are unable to meet the photogrammetric processing requirements of massive planetary remote sensing images. To address these issues, we first proposed an efficient control network construction framework based on approximate orthoimage matching and hash quick search. Next, to effectively reduce the redundant control points in the control network and decrease the computation time required for bundle adjustment, we then proposed a control network-thinning algorithm based on a K-D tree fast search. Finally, we developed an automatic detection method based on ray tracing for identifying invalid control points in the control network. To validate the proposed methods, we conducted photogrammetric processing experiments using both the Lunar Reconnaissance Orbiter (LRO) narrow-angle camera (NAC) images and the Origins Spectral Interpretation Resource Identification Security Regolith Explorer (OSIRIS-REx) PolyCam images; we then compared the results with those derived from the famous open-source planetary photogrammetric software, the United States Geological Survey (USGS) Integrated Software for Imagers and Spectrometers (ISIS) version 8.0.0. The experimental results demonstrate that the proposed methods significantly improve the efficiency and quality of constructing control networks for large-scale planetary images. For thousands of planetary images, we were able to speed up the generation and editing of the control network by more than two orders of magnitude.

Determining match pairs among disordered images is a prerequisite for constructing tie points and control networks for the bundle adjustment of planetary remote sensing images. The efficiency and accuracy of determining image match pairs greatly affect the performance of generating planetary mapping products and three-dimensional models. However, existing methods for determining image match pairs in the field of planetary mapping generally use exhaustive methods (such as the USGS ISIS software), which are extremely inefficient when processing a large number of images. This paper proposes an efficient method of image overlapping relationship analysis based on the spatial index of KD tree fast search for disordered and large-scale asteroid images. The algorithm first calculates the three-dimensional coordinates of the image center point, constructs a KD tree data structure based on the three-dimensional coordinates of each image center, and then uses the KD tree fast search algorithm to quickly find the image overlapping relationships within the spec-ified spatial range. For situations where the entire celestial body only occupies a local area of the remote sensing image during the approach stage of asteroids, the center of the effective pixels in the image is calculated by the ray intersection of sampling points. In this study, the image data from asteroid exploration missions such as Bennu, Vesta, and Ryugu are used for experiments, and the proposed image match pairs determination algorithm is systematically compared with the corre-sponding modules of USGS ISIS in order to evaluate its performance in terms of efficiency and accuracy. The results show that when processing more than a thousand images, the proposed method is two to three orders of magnitude higher in efficiency than the corresponding module in USGS ISIS while ensuring the correctness of the image overlapping relationships and the accuracy of bundle adjustment, with the advantage becoming clearer as the amount of data increases. At the same time, according to the obtained image match pairs, images that meet the requirements of Stereo Photoclinometry can be quickly and preferably selected, effectively improving the quality of 3D reconstruction models of asteroid images.

High-resolution and high-accuracy mapping products of the Lunar South Pole (LSP) will play a vital role in future lunar exploration missions. Existing lunar global mapping products cannot meet the needs of engineering tasks, such as landing site selection and rover trajectory planning, at the LSP. The Lunar Reconnaissance Orbiter (LRO)’s narrow-angle camera (NAC) can acquire submeter images and has returned a large amount of data covering the LSP. In this study, we combine stereo-photogrammetry and photoclinometry to generate high-resolution digital orthophoto maps (DOMs) and digital elevation models (DEMs) using LRO NAC images for a candidate landing site at the LSP. The special illumination and landscape characteristics of the LSP make the derivation of high-accuracy mapping products from orbiter images extremely difficult. We proposed an easy-to-implement shadow recognition and contrast stretching method based on the histograms of the LRO NAC images, which is beneficial for photogrammetric and photoclinometry processing. In order to automatically generate tie points, we designed an image matching method considering LRO NAC images’ features of long strips and large data volumes. The terrain and smoothness constraints were introduced into the cost function of photoclinometry adjustment, excluding pixels in shadow areas. We used 61 LRO NAC images to generate mapping products covering an area of 400 km². The spatial resolution of the generated DOMs was 1 m/pixel, and the grid spacing of the derived DEMs was 1 m (close to the spatial resolution of the original images). The generated DOMs achieved a relative accuracy of better than 1 pixel. The geometric accuracy of the DEM derived from photoclinometry was consistent with the lunar orbiter laser altimeter (LOLA) DEM with a root mean square error of 0.97 m and an average error of 0.17 m.