Impact crater recognition methods:A review

作     者：Dong CHEN Fan HU Liqiang ZHANG Yunzhao WU Jianli DU Jiju PEETHAMBARAN 

作者机构：College of Civil EngineeringNanjing Forestry UniversityNanjing 210037China Department of GeographyState Key Laboratory of Remote Sensing ScienceFaculty of Geographical ScienceBeijing Normal UniversityBeijing 100875China Purple Mountain ObservatoryChinese Academy of SciencesNanjing 210023China Key Laboratory of Space Object and Debris ObservationChinese Academy of SciencesNanjing 210023China Department of Mathematics and Computing ScienceSaint Mary’s UniversityHalifaxNS B3H 3C3Canada 

出 版 物：《Science China Earth Sciences》 (中国科学（地球科学英文版）)

年 卷 期：2024年第67卷第6期

页      面：1719-1742页

核心收录：

学科分类：07[理学] 0708[理学-地球物理学] 070401[理学-天体物理] 0704[理学-天文学] 

基　　金：supported by the National Natural Science Foundation of China(Grant Nos.41925006 12003075 42371383 and 42271450)。 

主　　题：Terrestrial planets Deep space exploration Impact crater Impact crater recognition Deep learning 

摘      要：Impact craters are formed due to the high-speed collisions between small to medium-sized celestial bodies.Impact is the most significant driving force in the evolution of celestial bodies,and the impact craters provide crucial insights into the formation,evolution,and impact history of celestial bodies.In this paper,we present a detailed review of the characteristics of impact craters,impact crater remote sensing data,recognition algorithms,and applications related to impact craters.We first provide a detailed description of the geometric texture,illumination,and morphology characteristics observed in remote sensing data of craters.Then we summarize the remote sensing data and cataloging databases for the four terrestrial planets(i.e.,the Moon,Mars,Mercury,and Venus),as well as the impact craters on Ceres.Subsequently,we study the advancement achieved in the traditional methods,machine learning methods,and deep learning methods applied to the classification,segmentation,and recognition of impact craters.Furthermore,based on the analysis results,we discuss the existing challenges in impact crater recognition and suggest some solutions.Finally,we explore the implementation of impact crater detection algorithms and provide a forward-looking perspective.