CaSSIS-based stereo products for Mars after three years in orbit
Graphical abstract
Introduction
The Colour and Stereo Surface Imaging System (CaSSIS) (Thomas et al., 2017) with its stereo capabilities was conceived for enhancing our knowledge of the surface of Mars by extending and complementing the observations of previous instruments such as the High Resolution Imaging Science Experiment (HiRISE) (McEwen et al., 2007) which is currently operating on board NASA's Mars Reconnaissance Orbiter (MRO). Even before, other systems have imaged the surface of Mars in stereo mode. The High Resolution Stereo Camera (HRSC) (Neukum and Jaumann, 2004) (Jaumann et al., 2007); of the Mars Express mission was one of the first stereo camera designed to derive digital terrain models (DTMs) and the corresponding orthoimages on its standard operation mode with resolution better than 20 m/pixel. Furthermore, the Context Camera (CTX (Malin et al., 2007)), for other MRO observations, is able to acquire stereoscopic image pairs with imaging scale similar to CaSSIS (∼6 m/pixel) operating in push broom mode. Both HiRISE and CTX acquire stereo pairs thanks to the ability of MRO to point off-nadir such that the camera can be pointed at specific targets.
CaSSIS adopts a single focal plane with four fixed colour filters and is able to acquire along-track stereo images thanks to a rotation mechanism. In the push-frame approach, the 2D image (hereafter called “framelet”) is acquired, then buffered, read while the spacecraft moves and transferred to the proximity electronics before the next framelet is acquired, considering a sufficient surface overlap with the previous one. In the moving of the spacecraft along the orbit, CaSSIS acquires push-frame images to build up a full image swath (first of the pair) before rotating by 180°, thanks to a rotational stage, to acquire the second image of the pair. The telescope, furthermore, is mounted tilted by 10° with respect to the mechanical axis of symmetry of the bearing (nadir direction) guaranteeing a stereo convergence angle of 22.4° for a circular orbit at a height of around 400 km above Mars surface.
CaSSIS is the most recent example of a stereo camera employed in the photogrammetric mapping for planetary exploration. The processing techniques described here, applied to the imaging data, allow us to derive high-level stereo products such as the high-resolution DTMs, the corresponding orthoimages and anaglyphs. The expectations on the CaSSIS data are to provide stereo coverage over the surface equivalent to ∼1–2% of the planet for every Martian year in orbit, complementing the other stereo data in terms of high resolution, colour and 3D mapping. The availability of three-dimensional data of a planetary surface enables the quantitative morphological analysis of any particular features, such as impact crater interiors and their ejecta blankets, the volcanic domes, mountain chains or hills, mounds, drainage divides, as well as incised or inverted riverbeds and deltas.
More specifically, CaSSIS DTMs and orthoimages have been already used to investigate the stratigraphy of the South Polar Layered Deposits (Becerra et al., 2019). In addition, they have been used to perform the topographic correction on CaSSIS orthoimages (i.e., modeling and removal of brightness differences induced by topography), allowing to obtain corrected images that will be key for future photometric analyses of Maritan surface features (Munaretto et al., 2021).
The ability of CaSSIS to acquire along-track stereo, near-simultaneous stereo pairs avoids illumination differences between images and offers benefits for the DTMs and anaglyphs production.
This is especially beneficial since the orbit of TGO precesses through all local times of the day (Beta angle), and revisit times are much farther apart (depending on latitude) than those of MRO, which is in a Sun-synchronous orbit, but acquires stereo pairs on different orbits. In this paper, a description of these CaSSIS stereo products is proposed with particular attention to the effect of input parameters on the quality and precision of the DTMs. In the photogrammetric processing section, the general procedure is described and a series of different processing parameters is applied to test several strategies driven by the scientific investigation scope. The variability of the input data (image content and textures related to the illumination conditions and morphologic characteristics) suggests the application of flexible processing parameters for the derivation of highly accurate DTMs. The quality of the products has been assessed in terms of internal accuracy in terms of completeness of the surface, details of the reconstructed surface shape and morphologic consistency. The results are compared with DTMs provided by other Mars imagers (HiRISE and CTX). In order to extract robust statistics and to be able to derive considerations from the comparison a co-registration process between the datasets has been fundamental.
The paper is organized as follows. Section 2 describes the photogrammetric processing implemented in the OAPD-INAF pipeline (3DPD software). In Section 3, the CaSSIS stereogrammetric products are described and in Section 4 the anaglyph processing is presented. Section 5 elaborates the methodology and describes the experimental dataset. In Section 6, the performance analysis of the photogrammetric products is reported in details, and is also supported also by qualitative analysis. A final discussion and conclusions are provided in Section 7.
Section snippets
Photogrammetric processing
The Observatory of Padova team (OAPD-INAF) 3DPD pipeline (Simioni et al., 2021) is currently generating stereo data products (DTMs and orthoimages) from the CaSSIS images. All the data are delivered to a dedicated repository accessible through a web-interface (https://cassis.oapd.inaf.it/archive) (Cremonese et al., 2018) that is designed to manage the distribution of the DTMs (and connected ancillary products) within the CaSSIS team. The system provides on-line visualization and permits the
CaSSIS stereo products
The OAPD-INAF team has produced 263 DTMs so far on a total of more than 1700 stereo pairs acquired by the instrument (through September 2020). The locations of the targets acquired and the corresponding DTMs delivered are reported in Fig. 2 and Fig. 3, respectively.
The repository includes the DTMs that are 32-bit signed, produced in Equirectangular projection equally sampled in planetographic Latitude, with elevations referenced to the MOLA datum (version megR_hb - https://pds-geosci
Anaglyph processing
Another CaSSIS stereo product is represented by Anaglyphs, produced by the University of Arizona team. These products, which can be viewed in 3D with standard red-blue glasses, are assembled by placing the image from one half of a stereo pair into the red channel (for the left eye) and the image from the other half of the stereo pair in the green and blue channels (for the right eye) of an RGB image. Anaglyphs can aid in the qualitative analysis of complex surfaces with a minimized amount of
Methodology and experimental data
The relative quality of the output DTM largely depends on the dense matching algorithms applied, which are dependent on the quality of the images themselves both in terms of radiometric content (Pommerol, 2021) and geometric properties (Tulyakov et al., 2018); while the issues concerning absolute accuracy are not considered in this context. Since the quality of the DTMs affects the accuracy in computing multiple morphometric indexes (i.e., mean gradients, sharpness, curvature maps and slope
DTMs evaluation method and Co-registration
Regarding the DTM evaluation, an ideal reference dataset should be available in order to investigate the quality measures. In particular, the reference DTM should be independent with respect to the DTM to be assessed and preferably with higher precision.
Since in the planetary context, the ground control points are not available as on the Earth, the quality of the 3D reconstruction of the planetary surfaces must analyze internal consistency and comparisons with complementary remote sensing data (
Summary and conclusions
This work provides a review of the CaSSIS stereo products supported by a description of the applied methods and examines the correlation between the quality of the images in terms of image content and the topographic characteristics with the matching performance that directly affects the quality of DTMs.
Excluding from the discussion the “Hebes Chasma” area that misses the HiRISE reference model (and the derived RMSE of the discrepancies), even though the aim of the paper does not foresee going
Author statement
C. Re: Conceptualization, Methodology, Software, Formal analysis, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision.
A. Fennema: Software, Data Curation, Metodology, Writing - Review & Editing.
E. Simioni: Software, Data Curation, Visualization, Supervision.
S. Sutton: Review & Editing, Supervision.
D. Mège: Writing, Metodology, Visualization.
K. Gwinner: Supervision.
M. Józefowicz: Resources.
G. Munaretto: Resources.
M. Pajola: Editing.
A. Petrella: Data
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
CaSSIS is a project of the University of Bern and funded through the Swiss Space Office via ESA's PRODEX programme. The instrument hardware development was also supported by the Italian Space Agency (ASI) (ASI-INAF agreement no. 2020-17-HH.0), INAF/Astronomical Observatory of Padova, and the Space Research Center (CBK) in Warsaw. Support from SGF (Budapest), the University of Arizona (Lunar and Planetary Laboratory) and NASA are also gratefully acknowledged. Operations support from the UK Space
References (51)
Ancillary data services of NASA's navigation and ancillary information facility
Planet. Space Sci.
(1996)- et al.
Speeded-up robust features (SURF)
Comput. Vis. Image Understand.
(2008) - et al.
Clementine imagery: selenographic coverage for cartographic and scientific use
Planet. Space Sci.
(1996) - et al.
Color imaging of Mars by the high resolution imaging science experiment (HiRISE)
Icarus
(2010) - et al.
The high-resolution stereo camera (HRSC) experiment on Mars Express: instrument aspects and experiment conduct from interplanetary cruise through the nominal mission
Planet. Space Sci.
(2007) - et al.
Modeling the formation of bright slope deposits associated with gullies in Hale Crater, Mars: implications for recent liquid water
Icarus
(2010) - et al.
The high resolution imaging science experiment (HiRISE) during MRO's primary science phase (PSP)
Icarus
(2010) Recent water release in the Tharsis region of Mars
Icarus
(1990)- et al.
Topographic correction of HiRISE and CaSSIS images: validation and application to color observations of Martian albedo features
Planet. Space Sci.
(2021) - et al.
Exploration and analysis of planetary shape and topography using stereophotogrammetry
Encyc. Solar Syst.
(2014)