Automatic extraction of lunar impact craters from Chang’E-1 satellite photographs

Abstract

The distribution characteristics of the impact craters can provide a large amount of information on impact history and the lunar evolution process. In this research, based on the digital elevation model (DEM) data originating from Change'E-1 CCD stereo camera, three automatic extraction methods for the impact craters are implemented in two research areas: direct extraction from flooded DEM data (the Flooded method), object-oriented extraction from DEM data by using ENVI ZOOM function (the Object-Oriented method) and novel object-oriented extraction from flooded DEM data (the Flooded Object-Oriented method). Accuracy assessment, extracted degree computation, cumulative frequency analysis, shape and age analysis of the extracted craters combined display the following results. (1) The Flooded Object-Oriented method yields better accuracy than the other two methods in the two research areas; the extraction result of the Flooded method offers the similar accuracy to that of the Object-Oriented method. (2) The cumulative frequency curves for the extracted craters and the confirmed craters share a similar change trajectory. (3) The number of the impact craters extracted by the three methods in the Imbrian period is the largest and is of various types; as to their age earlier than Imbrain, it is difficult to extract because they could have been destroyed. As one of the most typical geomorphologic units, the impact craters are widely distributed on the lunar surface. The distribution characteristics of the impact craters provide a large amount of information on impact history and the lunar evolution process. They can be used to estimate lunar surface geology and relative age [1,2], and they also help to explain crust structures of the lunar surface [3]. In order to acquire the information, a mass of craters need to be extracted from images. Manual identification of craters is very time consuming, so researchers presented many automatic extraction methods, such as an algorithm which was able to automatically detect and classify 80% of craters without parameter tuning [4] and a method that enabled a fully automatic detection of 70% of sub-km craters in large panchromatic images [5]. Their extraction work used topographic and spectral information from lunar photos and images like Apollo data, Clementine UV-Vis is multi-spectral images and SELENE data. There was also much work [6][7][8]

Automatic detection of complex archaeological grazing structures using airborne laser scanning data

“…The method is then based on a geometrical template, defined from the knowledge of the morphometric characteristics of a targeted archaeological feature. The main problem is that the automatic detection is mostly limited to simple and isolated structures (Kim et al, 2005;Krøgli et al, 2007;Rack et al, 2005;Wan et al,2012). The aim of this paper is to propose a combination between pixelbased classification and template matching, in order to detect the elementary parts of complex archaeological structures.…”

A Bayesian Network Framework for Automatic Detection Of Lunar Impact Craters Based on Optical Images and Dem Data

“…Kim et al 2005 andMichael et al 2003 applied slope measurements to the identification of the craters, where they considered the area with a high slope change to be the edge of the crater, which successfully identified 75% of the craters with a diameter greater than 10 km. (Wan et al 2012) were implemented for automatic extraction of the impact craters based on based on the DEM data originating from Change'E-1 CCD stereo camera. ).…”

Shape characteristics-based extraction of lunar impact craters: using DEM from the Chang'E-1 satellite as a data source

“…The basin filling algorithm, which is employed to pre-process data before hydrological analysis, has been widely used for morphologic and drainage research (Wan et al 2012). It is designed to prevent the jam phenomenon when water passes by the basin.…”

“…Information of the structure and morphological characteristics of lunar impact craters in DEM data are more abundant and can quantitatively describe the topographic structure and the spatial pattern of craters. It will be more suitable for the analysis of impact crater morphology and extraction and can better solve the extraction problem caused by interactions among different impact craters (Kim and Muller 2003;Degirmenci and Ashyralyyev 2010;Wan et al 2012).…”