Erosion rate of the north polar scarps of Mars based on automated detection of block falls in HiRISE images

We automatically detected ice block falls in high resolution satellite images of the north polar region of Mars, with the aim of understanding their role in erosion processes and the morphological evolution of the region. Our change detection method is based on the combination of Binary Large OBjects (BLOB)detection and machine learning and we aim to extend our investigation to all steep scarps of the margin of the north polar ice cap where suitable high resolution data exist.