Automatic Recognition of Impact Craters on the Martian Surface from DEM and Images - Planetary Exploration and Science: Recent Results and Advances - page 97

Chapter 6
Automatic Recognition of Impact Craters
on the Martian Surface from DEM and Images

Tengyu Zhang and Shuanggen Jin

Abstract Impact craters are the most outstanding and attractive geomorphological
features on the surface of the planets, showing variety and complexity of the surface
morphology. The accurate recognition of impact craters on Mars is very useful to
analyze and understand the relative dating of Martian surface. In this chapter, four
crater-detection methods have been presented and discussed with various extent of
discrimination ability on Martian images or topography data. The modified ad boost-
ing approach demonstrates the best performance in classification of craters, while
the algorithms based on topography data have low efficiency in automatic detection.
Comparing to previous solutions, the modified ad boosting method has greatly
improved the detecting performance of the algorithm and reduced detection time.

Keywords Automatic detection ￿ Craters ￿ DEM ￿ Images

6.1 Introduction

Impact craters are natural probes of target surface properties. Recognitions of
impact craters have revealed lots of findings (Cheng et al. 2004 ; Johnson et al.
2005 ; Kim et al. 2005 ;Sawabeetal. 2006 ). Surface dating is one of the most
common applications by the analysis of the size and distribution frequency of
craters, indicating that a catalogue with the size and distribution is very useful in
combination with DEM (Digital Elevation Model) data (Crater Analysis Technique
Working Group 1979 ). The shapes of craters on Mars have lots of different types
according to the interior morphology (central peaks, peak rings, central pits, wall
terraces) and ejecta structures (pedestal, pancake, rampart, lobate, fluidized, radial