A Risk Assessment of Potential Landing Zones on Mars Using Gis, Machine Learning, and Stereo-Photogrammetry

With more nations gaining the capability to land spacecraft on Mars, a need is developing to more efficiently produce risk assessments for candidate landing zones. Currently, the only two cameras capable of sub-meter resolution as necessary to properly perform such risk assessments are HiRISE and HiRIC and their time is tightly budgeted. A preliminary risk assessment by a lower resolution camera with wide global coverage would lower the risk of wasting valuable satellite time taking imagery of a site that turns out to be too feature-rich to allow for a safe landing. This study fills that niche by developing a workflow that uses CTX imagery to perform a first-order risk assessment by combining image classification with a DTM. Hazards identified by machine learning can be combined with a slope map to produce a holistic risk assessment of any proposed landing zone with CTX stereo-coverage.