Crater Detection Using Unsupervised Algorithms and Convolutional Neural Networks

被引:28
作者
Emami, Ebrahim [1 ]
Ahmad, Touqeer [1 ]
Bebis, George [1 ]
Nefian, Ara [2 ]
Fong, Terry [2 ]
机构
[1] Univ Nevada, Dept Comp Sci & Engn, Reno, NV 89577 USA
[2] NASA, Intelligent Robot Grp, Ames Res Ctr, Moffett Field, CA 94035 USA
来源
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING | 2019年 / 57卷 / 08期
基金
美国国家航空航天局;
关键词
Convex grouping; convolutional neural networks (CNNs); crater detection; MARTIAN IMPACT CRATERS; AUTOMATIC DETECTION; OPTICAL-IMAGES; TOPOGRAPHY; CATALOG; LUNAR;
D O I
10.1109/TGRS.2019.2899122
中图分类号
P3 [地球物理学]; P59 [地球化学];
学科分类号
0708 ; 070902 ;
摘要
Craters are among the most abundant features on the surface of many planets with great importance for planetary scientists. They reveal chronology information about planets and may be used for autonomous spacecraft navigation and landing. Although numerous research efforts have been carried out in the field of crater detection, existing crater detection algorithms (CDAs) are only helpful in a limited number of applications. A promising crater detection approach involves two main steps: 1) hypothesis generation (HG) and 2) hypothesis verification (HV). During HG, potential crater locations are detected. The validity of the hypothesized crater locations is then tested in a HV step. In this context, we discuss some commonly used algorithms for HG such as highlight-shadow region detection and Hough transform as well as our novel and enhanced algorithms based on interest point detection and convex grouping. A key objective of this paper is to analyze their performance while paying special attention to how they affect the accuracy of the verification step. To deal with different size craters, we focus on multiscale HG. For HV, we have chosen convolutional neural networks which have recently achieved state-of-the-art performance in many computer vision applications. Due to the variation of test sets in the literature, it is often challenging to compare the performance of different CDAs in a fair way. In this paper, we present a comprehensive performance evaluation and comparison of CDAs. Each algorithm has been trained/tested using common data sets generated by a systematic approach.
引用
收藏
页码:5373 / 5383
页数:11
相关论文
共 35 条
[1]   Impact crater recognition on mars based on a probability volume created by template matching [J].
Bandeira, Lourenco ;
Saraiva, Jose ;
Pina, Pedro .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2007, 45 (12) :4008-4015
[2]   Constraining geologic properties and processes through the use of impact craters [J].
Barlow, Nadine G. .
GEOMORPHOLOGY, 2015, 240 :18-33
[3]   Machine detection of Martian impact craters from digital topography data [J].
Bue, Brian D. ;
Stepinski, Tomasz F. .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2007, 45 (01) :265-274
[4]  
Challenge for NASA Tournament Lab, CRAT DET
[5]   The Complexity of DNF of Parities [J].
Cohen, Gil ;
Shinkar, Igor .
ITCS'16: PROCEEDINGS OF THE 2016 ACM CONFERENCE ON INNOVATIONS IN THEORETICAL COMPUTER SCIENCE, 2016, :47-58
[6]   Novel approach of crater detection by crater candidate region selection and matrix-pattern-oriented least squares support vector machine [J].
Ding Meng ;
Cao Yunfeng ;
Wu Qingxian .
CHINESE JOURNAL OF AERONAUTICS, 2013, 26 (02) :385-393
[7]   On Crater Verification Using Mislocalized Crater Regions [J].
Emami, Ebrahim ;
Bebis, George ;
Nefian, Ara ;
Fong, Terry .
2017 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2017), 2017, :1098-1104
[8]   Automatic Crater Detection Using Convex Grouping and Convolutional Neural Networks [J].
Emami, Ebrahim ;
Bebis, George ;
Nefian, Ara ;
Fong, Terry .
ADVANCES IN VISUAL COMPUTING, PT II (ISVC 2015), 2015, 9475 :213-224
[9]  
Galloway MJ, 2014, IEEE IMAGE PROC, P1579, DOI 10.1109/ICIP.2014.7025316
[10]   Cratering chronology and the evolution of Mars [J].
Hartmann, WK ;
Neukum, G .
SPACE SCIENCE REVIEWS, 2001, 96 (1-4) :165-194