An algorithm was developed that automatically processes images captured by the HiRISE camera on the Mars Reconnaissance Orbiter to identify and locate the presence of purgatoids in monochrome images.
Purgatoids are eolian bedforms that present significant mobility risks to wheeled vehicles on Mars. The algorithm employs a series of filters and image processing steps to enhance the images and isolate each purgatoid. Classification is finally achieved by applying a series of selection criteria.
This work was done by Paolo Bellutta of Caltech and Steven B. Wood of the University of Nevada, Reno for NASA’s Jet Propulsion Laboratory.