REAL-TIME COMPILATION OF COMPOSITE ROVER/ORBITER ORTHOMOSAIC AND DIGITAL ELEVATION MODEL BASE MAPS OF ROVER TRAVERSES FOR GEOLOGIC MAPPING

We are assembling base maps of combined orthorectified orbiter and rover stereo images (ORRs) and digital elevation models (DEMs) in real time for the Opportunity and Curiosity rovers. The Navigation Camera (Navcam) panoramas from the rovers are high-pass filtered to correct for shading differences between adjacent stereo “wedges” to produce a seamless ORR. This mosaic is then combined with the HiRISE red channel grayscale image to reintroduce the meter-scale shading from the HiRISE ORR into the Navcam ORR. The two are then combined with the local HiRISE color data to produce a seamless ORR that preserves the centimeter-scale navcam resolution within a radius of 35 meters (Opportunity) or 45 meters (Curiosity) of the rovers’ end-of-drive positions. In localities where the rovers have conducted “walkabouts” – where the vehicle made several stops in areas of scientific interest that resulted in overlapping ORR/DEM coverage, multiple end of drive ORR/DEMs are combined to produce a more regional coverage map at 1cm/pixel (e.g., Spirit of Matijevic Hill and Marathon Valley by Opportunity; Yellowknife Bay and Bagnold Dunes by Curiosity).

Prior to the Curiosity landing, the science team divided the landing ellipse and the science target area south of the ellipse into 0.025° quadrangles that were then mapped by members of the team to develop a working geologic map of the terrains anticipated to be visited by the rover. Geologic mapping continues, including image data acquired by the rover as it traversed through successive quadrangles. We are just beginning to delineate similar size quadrangles for the Opportunity traverse, beginning with the current area of investigation in Marathon Valley, and extending both to the south along the planned Extended Mission 10 traverse route, and to the north along the traverse from Cape York along the Endeavour Crater rim.

We feel that the rover and HiRISE image products have greater utility for geologic mapping when combined in this manner than treated as separate perspective products. Instead of restricting mapping to a narrow “corridor” of very high resolution data acquired via the ground view, this product enables local and regional context inferences to be made, because image seams, shading differences, and elevation offsets are corrected as the merged products are created.