A Martian rover capable of analyzing images autonomously could traverse gre
ater path lengths and return data with greater scientific content. A more i
ntelligent rover could, for example, automatically select targets of intere
st (e.g., rocks, layers), return spectral or high-resolution image data of
these targets at the same time, remove less interesting or redundant parts
of images before transmitting them, and provide compact information or repr
esentations of its environment. Three prototype algorithms, a horizon detec
tor, a rock detector, and a layer detector have been developed and tested d
uring the 1999 Marsokhod rover field test in Silver Lake, California. The r
esults are encouraging and demonstrate the potential savings in time as wel
l as the potential increase in the amount of relevant science data returned
in each command cycle.