基于深度学习的大型陨石坑识别方法研究

doi:10.13700/j.bh.1001-5965.2019.0342

北京航空航天大学学报 ›› 2020, Vol. 46 ›› Issue (5): 994-1004.doi: 10.13700/j.bh.1001-5965.2019.0342

基于深度学习的大型陨石坑识别方法研究

郑磊¹, 胡维多¹, 刘畅²

1. 北京航空航天大学宇航学院, 北京 100083;
2. 中国科学院深圳先进技术研究院, 深圳 518055

收稿日期:2019-06-28 发布日期:2020-05-19
通讯作者: 胡维多 E-mail:08109@buaa.edu.cn
作者简介:郑磊男,硕士研究生。主要研究方向:基于深度学习的图像识别和目标检测方法;胡维多男,博士,教授,博士生导师。主要研究方向:轨道动力学、航天器姿态控制、图像处理;刘畅男,博士,副研究员。主要研究方向:图像分割算法、基于计算机视觉的目标三维重建、三维姿态估计与实时跟踪。
基金资助:
国家自然科学基金（61703017）

Large crater identification method based on deep learning

ZHENG Lei¹, HU Weiduo¹, LIU Chang²

1. School of Astronautics, Beihang University, Beijing 100083, China;
2. Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China

Received:2019-06-28 Published:2020-05-19

摘要/Abstract

摘要： 陨石坑是天体表面最为显著的地形特征，传统陨石坑识别方法主要是对小型陨石坑正负样本的二分类问题研究，且效率和精度均不高。以星体宏观视角下的大型陨石坑作为研究对象，结合图像处理和神经网络等方面的知识，创建了来自不同数据源的陨石坑样本数据库，研究了数据源对网络模型泛化能力的影响，提出了一种效率更高的陨石坑多分类识别方法。在非极大值抑制（NMS）算法基础上，提出了一种精度更高的陨石坑检测算法。经过参数优化和实验验证，构建的基于深度学习的多尺度多分类陨石坑自动识别网络框架取得了较高的准确率，在同源验证集上识别率可达0.985，在异源验证集上识别率可达0.863，并且有效改善了目标检测时检测框冗余及误检测的问题。

关键词: 深度学习, 卷积神经网络, 陨石坑识别, 非极大值抑制(NMS)算法, 目标检测

Abstract: Craters are the most significant topographic features on the surface of celestial bodies. The traditional method of craters identification is mainly to study the dichotomy of positive and negative samples of small craters, with low efficiency and accuracy. This paper takes large craters under the macroscopic view of the planet as the research object, combines the knowledge of digital image processing and neural network, creates a crater sample library of different data sources to study the influence of data source on network model generalization ability, and proposes a more efficient crater multi-classification identification method. Based on the Non-Maximum Suppression (NMS) algorithm, a higher precision crater detection algorithm is proposed. Through parameter optimization and experimental verification, the multi-scale and multi-classification craters automatic recognition network framework based on deep learning constructed in this paper achieves a high accuracy rate, with the recognition rate up to 0.985 on homologous verification set and 0.863 on heterogeneous verification set, and effectively improves the redundancy of detection box and false detection in target detection.

Key words: deep learning, convolutional neural network, crater identification, Non-Maximum Suppression (NMS) algorithm, object detection

中图分类号:

郑磊, 胡维多, 刘畅. 基于深度学习的大型陨石坑识别方法研究[J]. 北京航空航天大学学报, 2020, 46(5): 994-1004.

ZHENG Lei, HU Weiduo, LIU Chang. Large crater identification method based on deep learning[J]. Journal of Beijing University of Aeronautics and Astronautics, 2020, 46(5): 994-1004.

参考文献

[1] GROUP C A T W,ARVIDSON R E,BOYCE J,et al.Standard techniques for presentation and analysis of crater size-frequency data[J].Icarus,1979,37(2):467-474.
[2] PALAFOX L F,HAMILTON C W,SCHEIDT S P,et al.Automated detection of geological landforms on Mars using convolutional neural networks[J].Computers & Geosciences,2017,101:48-56.
[3] CHENG Y,JOHNSON A E,MATTHIES L H,et al.Optical landmark detection for spacecraft navigation[C]//Proceedings of the 13th Annual AAS/AIAA Space Flight Mechanics Meeting,2003: 1785-1803.
[4] SAWABE Y,MATSUNAGA T,ROKUGAWA S.Automated detection and classification of lunar craters using multiple approaches[J].Advances in Space Research,2006,37(1):21-27.
[5] KIM J R,MULLER J,VAN GASSELT S,et al.Automated crater detection,a new tool for Mars cartography and chronology[J].Photogrammetric Engineering & Remote Sensing,2005,71(10):1205-1217.
[6] 冯军华, 崔祜涛,崔平远,等.行星表面陨石坑检测与匹配方法[J].航空学报,2010,31(9)：1858-1863.FENG J H,CUI H T,CUI P Y,et al.Autonomous crater detection and matching on planetary surface[J].Acta Aeronautica et Astronautica Sinica,2010,31(9):1858-1863(in Chinese).
[7] DING M,CAO Y F,WU Q X.Method of passive image based crater autonomous detection[J].Chinese Journal of Aeronautics,2009,22(3):301-306.
[8] CRACKNELL M J,READING A M.Geological mapping using remote sensing data: A comparison of five machine learning algorithms,their response to variations in the spatial distribution of training data and the use of explicit spatial information[J].Computers & Geosciences,2014,63:22-33.
[9] CHRISTOFF N,MANOLOVA A,JORDA L,et al.Morphological crater classification via convolutional neural network with application on MOLA data[C]//Advances in Neural Networks and Applications 2018,2018:1-5.
[10] BARATA T,ALVES E I,SARAIVA J,et al.Automatic recognition of impact craters on the surface of Mars[C]//International Conference Image Analysis and Recognition.Berlin：Springer,2004:489-496.
[11] BOUKERCHA A,AL-TAMEEMI A,GRUMPE A,et al.Automatic crater recognition using machine learning with different features and their combination[C]//Lunar and Planetary Science Conference,2014,45:2842.
[12] SILBURT A,ALI-DIB M,ZHU C,et al.Lunar crater identification via deep learning[J].Icarus,2019,317:27-38.
[13] WRIGHT E.SVS: Moon phase and libration,2018[EB/OL].(2019-01-28)[2019-06-12].https://svs.gsfc.nasa.gov/4604.
[14] HEAD J W,ADAMS J B,MCCORD T B,et al.Regional stratigraphy and geologic history of Mare Crisium[C]//Mare Crisium:The View From Luna 24,1978:43-74.
[15] WATTERS T R,KONOPLIV A S.The topography and gravity of Mare Serenitatis:Implications for subsidence of the mare surface[J].Planetary and Space Science,2001,49(7):743-748.
[16] SOLOMON S C,HEAD J W.Vertical movement in mare basins:Relation to mare emplacement,basin tectonics,and lunar thermal history[J].Journal of Geophysical Research: Solid Earth,1979,84(B4):1667-1682.
[17] SCHULTZ P H,STAID M I,PIETERS C M.Lunar activity from recent gas release[J].Nature,2006,444(7116):184-186.
[18] KRVGER T,VAN DER BOGERT C H,HIESINGER H.Geomorphologic mapping of the lunar crater Tycho and its impact melt deposits[J].Icarus,2016,273:164-181.
[19] ZHANG K,ZUO W,CHEN Y,et al.Beyond a Gaussian denoiser:Residual learning of deep CNN for image denoising[J].IEEE Transactions on Image Processing,2017,26(7):3142-3155.
[20] DONG C,LOY C C,HE K,et al.Image super-resolution using deep convolutional networks[J].IEEE transactions on Pattern Analysis and Machine Intelligence,2015,38(2):295-307.
[21] LONG J,SHELHAMER E,DARRELL T.Fully convolutional networks for semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE Press,2015:3431-3440.
[22] MALTEZOS E,DOULAMIS N,DOULAMIS A,et al.Deep convolutional neural networks for building extraction from orthoimages and dense image matching point clouds[J].Journal of Applied Remote Sensing,2017,11(4):42620.
[23] ZHANG N, DONAHUE J,GIRSHICK R,et al.Part-based R-CNNs for fine-grained category detection[C]//European Conference on Computer Vision.Berlin:Springer,2014:834-849.
[24] 黄洁,姜志国,张浩鹏,等.基于卷积神经网络的遥感图像舰船目标检测[J]北京航空航天大学学报,2017,43(9):1841-1848.HUAGN J,JIANG Z G,ZHANG H P,et al.Ship object detection in remote sensing images using convolutional neural networks[J].Journal of Beijing University of Aeronautics and Astronautics,2017,43(9):1841-1848(in Chinese).
[25] JIAO L,LIANG M,CHEN H,et al.Deep fully convolutional network-based spatial distribution prediction for hyperspectral image classification[J].IEEE Transactions on Geoscience and Remote Sensing,2017,55(10):5585-5599.
[26] 韩京冶,许福,陈志泊,等.一种基于深度学习的交互式电话号码识别方法[J].北京航空航天大学学报,2018,44(5):1074-1080.HAN J Y,XU F,CHEN Z B,et al.A deep learning based interactive recognition method for telephone numbers[J].Journal of Beijing University of Aeronautics and Astronautics,2018,44(5):1074-1080(in Chinese).
[27] KRIZHEVSKY A, SUTSKEVER I,HINTON G E.Imagenet classification with deep convolutional neural networks[C]//Advances in Neural Information Processing Systems,2012: 1097-1105.
[28] RIEDMILLER M,BRAUN H.A direct adaptive method for faster backpropagation learning: The RPROP algorithm[C]//Proceedings of the IEEE International Conference on Neural Networks.Piscataway:IEEE Press,1993:586-591.
[29] ZEILER M D,FERGUS R.Visualizing and understanding convolutional networks[C]//European Conference on Computer Vision.Berlin:Springer,2014:818-833.
[30] ZEILER M D.ADADELTA:An adaptive learning rate method[EB/OL].(2012-12-22)[2019-06-20].https://arxiv.org/abs/1212.5701.
[31] KINGMA D P,BA J.Adam:A method for stochastic optimization[EB/OL].(2014-12-22)[2019-06-20].https://arxiv.org/abs/1412.6980.
[32] TIELEMAN T, HINTON G.RMSPROP:Divide the gradient by a running average of its recent magnitude[J].Neural Networks for Machine Learning,2012,4:26-30.

基于深度学习的大型陨石坑识别方法研究

Large crater identification method based on deep learning

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	边亮, 罗霄阳, 李硕. 基于深度学习的图像拼接篡改检测[J]. 北京航空航天大学学报, 2020, 46(5): 1039-1044.
[2]	贾光辉, 于云瑞, 王丹. 卷积神经网络求解有限元单元刚度矩阵[J]. 北京航空航天大学学报, 2020, 46(3): 481-487.
[3]	陈映雪, 丁文锐, 李红光, 王蒙, 王旭. 基于视频帧间运动估计的无人机图像车辆检测[J]. 北京航空航天大学学报, 2020, 46(3): 634-642.
[4]	杨思晨, 王华锋, 王月海, 李锦涛, 王赟豪. 深度学习机制与小波融合的超分辨率重建算法[J]. 北京航空航天大学学报, 2020, 46(1): 189-197.
[5]	孙宇航, 曾国奇, 刘春辉, 张多纳. 基于深度学习的无人机数据链信噪比估计算法[J]. 北京航空航天大学学报, 2019, 45(9): 1855-1863.
[6]	厉铮泽, 杨小远, 朱日东, 王敬凯. 基于深度卷积的残差三生网络研究与应用[J]. 北京航空航天大学学报, 2019, 45(9): 1864-1873.
[7]	张克明, 蔡远文, 任元. 基于生成对抗网络的航天异常事件检测方法[J]. 北京航空航天大学学报, 2019, 45(7): 1329-1336.
[8]	张鹏, 张嘉峰, 刘涛. 基于相干度优化的极化顺轨干涉SAR慢小目标CFAR检测[J]. 北京航空航天大学学报, 2019, 45(3): 575-587.
[9]	魏宏喜, 张越. 基于生成对抗网络的零样本图像分类[J]. 北京航空航天大学学报, 2019, 45(12): 2345-2350.
[10]	张新峰, 郭宇桐, 蔡轶珩, 孙萌. 基于DCNN和全连接CRF的舌图像分割算法[J]. 北京航空航天大学学报, 2019, 45(12): 2364-2374.
[11]	翟宇轩, 刘怡桑, 徐艺文, 陈忠辉, 房颖, 赵铁松. 基于HTTP自适应流媒体传输的3D视频质量评价[J]. 北京航空航天大学学报, 2019, 45(12): 2456-2462.
[12]	杨彬, 李和平, 曾慧. 基于视频的三维人体姿态估计[J]. 北京航空航天大学学报, 2019, 45(12): 2463-2469.
[13]	张梦琴, 孟权令, 张维刚. 基于深度视觉语义嵌入的视频缩略图推荐[J]. 北京航空航天大学学报, 2019, 45(12): 2479-2486.
[14]	佟骏超, 吴熙林, 丁丹丹. 时空域上下文学习的视频多帧质量增强方法[J]. 北京航空航天大学学报, 2019, 45(12): 2506-2513.
[15]	贾瑞明, 刘圣杰, 李锦涛, 王赟豪, 潘海侠. 基于编解码双路卷积神经网络的视觉自定位方法[J]. 北京航空航天大学学报, 2019, 45(10): 1965-1972.