To explore high priority landing sites and to prepare for eventual human exploration, future Mars missions will involve rovers capable of traversing tens of kilometers. However, the current process by which scientists interact with a rover does not scale to such distances. Specifically, numerous command cycles are required to complete even simple tasks, such as, pointing the spectrometer at a variety of nearby rocks. In addition, the time required by scientists to interpret image data before new commands can be given and the limited amount of data that can be downlinked during a given command cycle constrain rover mobility and achievement of science goals. Experience with rover tests on Earth supports these concerns. As a result, traverses to science sites as identified in orbital images would require numerous science command cycles over a period of many weeks, months or even years, perhaps exceeding rover design life and other constraints. Autonomous onboard science analysis can address these problems in two ways. First, it will allow the rover to preferentially transmit "interesting" images, defined as those likely to have higher science content. Second, the rover will be able to anticipate future commands. For example, a rover might autonomously acquire and return spectra of "interesting" rocks along with a high-resolution image of those rocks in addition to returning the context images in which they were detected. Such approaches, coupled with appropriate navigational software, help to address both the data volume and command cycle bottlenecks that limit both rover mobility and science yield. We are developing fast, autonomous algorithms to enable such intelligent on-board decision making by spacecraft. Autonomous algorithms developed to date have the ability to identify rocks and layers in a scene, locate the horizon, and compress multi-spectral image data. We are currently investigating the possibility of reconstructing a 3D surface from a sequence of images acquired by a robotic arm camera. This would then allow the return of a single completely in focus image constructed only from those portions of individual images that lie within the camera's depth of field. Output from these algorithms could be used to autonomously obtain rock spectra, determine which images should be transmitted to the ground, or to aid in image compression. We will discuss these algorithms and their performance during a recent rover field test.


    Zugriff

    Zugriff über TIB

    Verfügbarkeit in meiner Bibliothek prüfen


    Exportieren, teilen und zitieren



    Titel :

    Autonomous Image Analysis for Future Mars Missions


    Beteiligte:
    Gulick, V. C. (Autor:in) / Morris, R. L. (Autor:in) / Ruzon, M. A. (Autor:in) / Bandari, E. (Autor:in) / Roush, T. L. (Autor:in)

    Erscheinungsdatum :

    1999-01-01


    Medientyp :

    Preprint


    Format :

    Keine Angabe


    Sprache :

    Englisch




    Autonomous Onboard Science Image Analysis for Future Mars Rover Missions

    Gulick, V. C. / Morris, R. L. / Ruzon, M. A. et al. | NTRS | 1999


    Potential future Mars missions

    Miller, Sylvia / Bell, Julia / Graf, James et al. | AIAA | 2000


    Future Missions to Mars

    Plescia, J. B. | NTRS | 1997


    Exobiology and Future Mars Missions

    Mckay, Christopher P. / Davis, Wanda, L. | NTRS | 1989


    Mars Characterization for Future Missions

    Bridger, Alison F. C. / Haberle, Robert M. | NTRS | 1999