月球阿波罗与嫦娥任务10个典型着陆区撞击坑数据库: 小撞击坑数据库构建与分布规律

刘方超; 张立; 郭弟均; 刘斌; 谢彬; 吕英波; 陈剑; 凌宗成

doi:10.11728/cjss2026.03.2025-0135

月球阿波罗与嫦娥任务10个典型着陆区撞击坑数据库: 小撞击坑数据库构建与分布规律

doi: 10.11728/cjss2026.03.2025-0135 cstr: 32142.14.cjss.2025-0135

刘方超¹,
张立²,
郭弟均^3, ,,
刘斌⁴,
谢彬⁴,
吕英波^1, ,,
陈剑¹,
凌宗成¹

1.
山东大学空间科学与技术学院　威海　264209
2.
山东大学低空科学与工程学院　威海　264209
3.
中国科学院国家空间科学中心太阳活动与空间天气全国重点实验室　北京　100190
4.
中国科学院空天信息创新研究院遥感与数字地球全国重点实验室　北京　100101

基金项目: 国家重点研发计划项目(2022YFF0711400)和国家自然科学基金项目(42472304)共同资助

详细信息

作者简介:

刘方超　男, 1992年10月出生于山东省潍坊市, 现为山东大学空间科学与技术学院, 博士研究生, 主要研究方向为行星科学与大数据

通讯作者:

郭弟均　男, 1988年9月出生于四川省巴中市, 现为中国科学院国家空间科学中心副研究员, 硕士生导师, 主要研究方向为月球与行星地质、行星遥感. E-mail: guodijun@nssc.ac.cn

吕英波　男, 1980年3月出生于山东省莱芜市, 现为山东大学空间科学与技术学院教授, 博士生导师, 主要研究方向为空间科学数据分析、深度学习与数据挖掘、深空资源勘探与利用、空间太阳能利用、材料模拟与工程仿真等相关的科学研究与技术攻关工作. E-mail: lyb@sdu.edu.cn

中图分类号: P691
计量
- 文章访问数: 337
- HTML全文浏览量: 71
- PDF下载量: 75
- 被引次数:
  0(来源:Crossref)
  
  0(来源:其他)
出版历程
- 收稿日期: 2025-07-31
- 修回日期: 2026-03-09
- 网络出版日期: 2026-03-12

Impact Crater Database of 10 Landing Regions from Apollo and Chang’E Missions: Construction and Distribution Patterns of Small Impact Crater Databases

1.
School of Space Science and Technology, Shandong University, Weihai 264209
2.
School of Low-altitude Science and Engineering, Shandong University, Weihai 264209
3.
State Key Laboratory of Solar Activity and Space Weather, National Space Science Center, Chinese Academy of Sciences, Beijing 100190
4.
State Key Laboratory of Remote Sensing and Digital Earth, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100101

摘要

摘要: 月球是撞击坑保存最完整的内太阳系天体, 拥有就位探测数据和返回样品的月球典型着陆区是认识撞击事件和撞击改造效应的重要研究对象. 其中小撞击坑对月壤的形成和演化具有重要影响, 但是现有的月球撞击坑数据库对直径<100 m的小尺度撞击坑覆盖不足. 因此, 本文筛选太阳入射角在50°～70°之间的月球高分辨率LROC NAC影像, 制作了覆盖10个着陆区20 km×20 km的镶嵌数据, 利用改进的YOLO11+SAHI深度学习模型对这10个着陆区中直径≥15 m的撞击坑进行自动提取, 经人工校验后建成了含359844条记录的撞击坑数据库. 系统构建了覆盖6个阿波罗任务与4个嫦娥任务共10个典型着陆区、直径≥15 m的撞击坑数据库, 且数据库中的撞击坑数据完整性优于现有研究. 进一步分析小尺寸撞击坑的密度分布与直径–频率分布特征. 本数据集可为月球地质年代标定、撞击通量演化、月表过程研究及样品分析提供高质量数据支持, 并为未来的撞击坑智能识别模型提供训练与验证基础.
- 小撞击坑 /
- 月球 /
- 着陆区 /
- 数据集 /
- 深度学习
Abstract: Among the celestial bodies within the solar system, the Moon maintains the most pristine conditions of impact craters. Some landing sites on the lunar surface have in-situ exploration data and experimental measurement results from the returned samples, providing distinguished meanings to understand the impact events and their reshaping effects on the lunar surface. Small craters play a significant role in the formation and evolution of lunar regolith. However, existing lunar impact crater databases lack comprehensive coverage of small-scale impact craters with diameters less than 100 m. Therefore, this study produced ten mosaic images of 20 km×20 km in width, using the high-resolution Lunar Reconnaissance Orbiter Camera Narrow Angle Camera images acquired under solar incidence angles between 50° and 70°. These images cover ten lunar landing sites, including six Apollo missions and four Chang’E missions. An improved YOLO11+SAHI deep learning model was then applied to automatically extract craters with diameters ≥15 m within these regions. After manual verification, a high-quality crater database containing 359844 records was constructed. Compared with existing datasets, the proposed database demonstrates superior crater completeness. Based on this database, the density distribution and diameter-frequency characteristics of small-sized impact craters were further calculated and analyzed. This dataset can provide robust support for studies of lunar geological chronology, impact flux evolution, surface process, and sample interpretation. In addition, it supplies a valuable training and validation resource for future artificial intelligent models of detecting the impact craters.
- Small-scale impact craters /
- Moon /
- Landing sites /
- Datasets /
- Deep learning
注释:

1) 1https://github.com/open-mmlab/labelbee

HTML全文

图 1 YOLO11+SAHI模型的网络结构

Figure 1. Architecture Diagram for YOLO11+SAHI Network

下载: 全尺寸图片幻灯片

图 2 选取的具有不同地形与不同光照条件的三个区域. GT表示真值, Pred表示预测值. (a)(d)表示反照率较高的高地区域, 整体撞击坑尺寸较小, (b)(e)表示反照率较低的月海区域, 撞击坑尺度为中等大小, (c)(f)表示反照率较低的月海区域, 存在一些大尺度的撞击坑

Figure 2. Schematic images with different illumination and terrain character. (a) (d) Represent high-albedo elevated regions with generally smaller crater sizes. (b) (e) Represent low-albedo lunar mare regions with medium-sized impact craters. (c) (f) Represent low-albedo lunar mare regions containing some large-scale impact craters

下载: 全尺寸图片幻灯片

图 3 10个着陆区的撞击坑标记 (绿色线表示研究区包含多个地质单元时着陆点所在地质单元的边界)

Figure 3. Impact crater maps density maps for ten landing regions (The green line indicates the boundary of the geological unit where the landing site is located when the study area contains multiple geological units)

下载: 全尺寸图片幻灯片

图 4 10个着陆区中的撞击坑大小频率累计分布 (垂直虚线表示完整性直径的位置)

Figure 4. Cumulative frequency distribution of crater sizes across ten landing regions (The vertical dashed lines indicate the positions of the integrity diameter)

下载: 全尺寸图片幻灯片

图 5 十个着陆区撞击坑核密度分布

Figure 5. Nuclear density distribution of impact craters in ten landing regions

下载: 全尺寸图片幻灯片

图 6 10个着陆区的撞击坑直径–频率分布的累积分布. (a) Apollo 11～17的累积分布, (b) Chang’E-3～6的累积分布

Figure 6. Cumulative distribution plots of crater diameter versus frequency for ten landing regions. (a) Cumulative plot for Apollo 11～17, (b) Cumulative plot for Chang’E-3～6

下载: 全尺寸图片幻灯片

图 7 10个着陆区的撞击坑直径–频率分布的R分布. (a)Apollo 11～17的R分布, (b) Chang’E-3～6的R分布

Figure 7. R-plots of crater diameter-frequency distributions for ten landing regions. (a) R-plots for Apollo 11～17, (b) R-plots for Chang’E-3～6

下载: 全尺寸图片幻灯片

表 1 构建训练集时选择的9个典型着陆区域的LRO NAC影像信息

Table 1. LRO NAC imagery information for the 9 landing regions where sample collection was conducted

着陆区	数据ID	分辨率/(m·pixel^–1)	太阳入射角/ (°)	相位角/ (°)
A11	M117338434 R	0.475	74.41	83.35
A12	M162466771 L	0.497	76.27	87.18
A14	M162426054 L	0.477	76.07	77.75
A15	M119829425 L	0.501	58.21	41.45
A16	M177535538 L	0.469	69.64	73.36
A17	M162107606 L	0.478	73.86	77.58
CE3	M1152001999 R	1.357	56.56	55.65
CE4	M1303619844 R	1.400	62.64	67.2
CE5	M1428381550 L+R	0.905	59.91	61.43

下载: 导出CSV

表 2 10个着陆区的镶嵌数据集范围

Table 2. Mosaic dataset coverage of ten landing regions

着陆区	纬度范围	经度范围
A11	0°20'40.75" －1°0'15.04"	23°8'35.54" －23°48'10.06"
A12	–3°20'32.49"－–2°40'58.20"	–23°45'7.82" －–23°5'30.18"
A14	–3°58'30.17" －–3°18'55.88"	–17°48'6.48" －–17°8'27.31"
A15	25°48'8.91" －26°27'43.20"	3°15'59.58" －4°0'4.21"
A16	–9°18'11.21" －–8°38'36.92"	15°10'2.43" －15°50'6.22"
A17	19°51'39.79" －20°31'14.08"	30°25'13.16" －31°7'22.93"
CE3	43°47'28.51" －44°27'2.80"	–19°58'15.06" －–19°3'7.74"
CE4	–45°47'14.84" －–45°7'40.55"	177°7'10.91" －178°3'36.00"
CE5	42°43'33.08" －43°23'7.37"	–52°22'8.01" －–51°27'58.71"
CE6	–41°58'4.97" －–41°18'30.68"	–154°25'36.71" －–153°32'39.61"

下载: 导出CSV

表 3 使用自然间断点分级法获得的10个着陆区各区间撞击坑直径区间划分以及区间内的桩径数量统计

Table 3. Statistics on the number of impact craters in each section of ten landing regions using Jenks natural breaks methods

着陆区	区间划分及撞击坑分区数量								总数量
A11	15～26	26～46	46～77	77～130	130～204	204～318	318～560	560～1534	39176
A11	29397	6163	1810	871	447	288	162	34	39176

A12	15～21	21～30	30～44	44～63	63～92	92～147	147～386	386～914	20598
A12	9393	4539	2187	1635	1434	922	447	35	20598

A14	15～21	21～32	32～49	49～76	76～141	141～272	272～472	472～1629	18564
A14	10132	4827	2053	851	426	151	60	63	18564

A15	15～24	24～36	36～54	54～83	83～128	128～206	206～437	437～834	27309
A15	17678	5486	2219	1034	499	277	104	10	27309

A16	15～27	27～44	44～72	72～121	121～197	197～320	320～615	615～1667	46889
A16	34608	8971	2196	712	226	83	48	43	46889

A17	15～22	22～32	32～50	50～79	79～124	124～201	201～355	355～1553	19066
A17	11628	4355	1808	525	367	237	102	43	19066

CE3	15～31	31～57	57～93	93～143	143～208	208～305	305～504	504～1216	31497
CE3	20297	5537	3310	1278	659	269	113	34	31497

CE4	15～26	26～43	43～68	68～98	98～137	137～247	247～515	515～1263	57234
CE4	33968	12353	6390	2634	1226	444	195	17	57234

CE5	15～36	36～60	60～96	96～140	140～201	201～297	297～447	447～966	56545
CE5	40650	8036	4980	1930	640	245	54	10	56545

CE6	15～30	30～47	47～70.8	71～97	97～133	133～198	198～338	338～853	42966
CE6	25458	11226	3018	1608	927	466	194	35	42966
注　第一列为着陆区名称, 每个着陆区对应数据的第一行为Jenks划分直径区间(单位: m), 第二行为各区间内统计的撞击坑数量, 最后一竖列为着陆区内撞击坑总数量.

下载: 导出CSV

表 4 参与训练的9个着陆点数据集详细信息及模型在其测试集上的准确率

Table 4. Detailed information on the nine landing site datasets used in training and the model’s accuracy on their test sets

着陆区	数据ID	标签数量	精确率	召回率	F1-score
A11	M117338434 R	1178	0.964	0.933	0.948
A12	M162466771 L	1131	0.938	0.902	0.919
A14	M162426054 L	1321	1	0.919	0.958
A15	M119829425 L	1282	0.993	0.939	0.965
A16	M177535538 L	3073	0.985	0.996	0.990
A17	M162107606 L	1823	1	0.964	0.982
CE3	M1152001999 R	1279	1	0.963	0.981
CE4	M1303619844 R	20785	0.993	0.915	0.952
CE5	M1428381550 L+R	5598	0.988	0.929	0.959
均值			0.985	0.94	0.962

下载: 导出CSV

表 5 10个着陆区检测到的撞击坑直径统计信息

Table 5. Statistical information on crater diameters detected across ten landing regions

着陆区	最小直径/m	最大直径/m	完整性直径/m	大于完整性直径的数量
A11	15	2239	22	14389
A12	15	1609	21	10023
A14	15	1641	21	7971
A15	15	3740	23	10808
A16	15	1817	22	19990
A17	15	2488	20	9184
CE3	15	1216	21	16975
CE4	15	1483	21	31203
CE5	15	965	26	26242
CE6	15	852	28	19618

下载: 导出CSV

表 6 10个着陆点所在地质单元的撞击坑密度和地质背景

Table 6. Impact crater density and geological background of the address units where the ten landing sites are located

着陆区	单个地质单元面积/km²	单个地质单元撞击坑数量	撞击坑密度/km^–2	年龄	地质单元	地质单元起伏情况
A11	400	39176	97.94	3.58～3.85 Ga^[28]	静海	平坦
A12	400	20595	51.49	3.15～3.22 Ga^[28]	风暴洋月海平原	平坦
A14	400	18564	46.41	3.77～3.85 Ga^[28]	弗拉·摩罗高地	崎岖
A15	247	20331	82.31	3.28～3.33 Ga^[28]	靠近盆缘高地的月海平原	平坦
A16	400	46889	117.22	3.77～3.85 Ga^[28]	笛卡尔高地	崎岖
A17	139	9560	68.78	3.5～3.85 Ga^[28]	靠近高地的月海平原	相对平坦
CE3	342	27685	80.95	2～3 Ga^[29]	虹湾月海平原	相对平坦
CE4	117	13653	116.69	2.95～3.67 Ga^[30]	被溅射物覆盖的月海平原	相对崎岖
CE5	400	56545	141.36	2.03 Ga^[31]	月海平原	相对平坦
CE6	400	42966	107.42	2.83 Ga^[32]	月海平原	平坦

下载: 导出CSV

参考文献(34)

[1]	XIAO Z Y, DI K C, XIE M G, et al. Impact flux on the moon[J]. Space: Science & Technology, 2024, 4(4): 0148
[2]	SHYLAJA B S. Determination of lunar surface ages from crater frequency—size distribution[J]. Journal of Earth System Science, 2005, 114(6): 609-612 doi: 10.1007/BF02715944
[3]	CHEN D, HU F, ZHANG L Q, et al. Impact crater recognition methods: a review[J]. Science China Earth Sciences, 2024, 67(6): 1719-1742 doi: 10.1007/s11430-023-1284-9
[4]	MICHAEL G G, NEUKUM G. Planetary surface dating from crater size-frequency distribution measurements: Partial resurfacing events and statistical age uncertainty[J]. Earth and Planetary Science Letters, 2010, 294(3-4): 223-229 doi: 10.1016/j.jpgl.2009.12.041
[5]	ROBBINS S J. A new global database of lunar impact craters >1-2 km: 1. Crater locations and sizes, comparisons with published databases, and global analysis[J]. Journal of Geophysical Research: Planets, 2019, 124(4): 871-892 doi: 10.1029/2018JE005592
[6]	WANG Y R, WU B, XUE H O, et al. An improved global catalog of lunar impact craters (≥1 km) with 3D morphometric information and updates on global crater analysis[J]. Journal of Geophysical Research: Planets, 2021, 126(9): e2020JE006728 doi: 10.1029/2020JE006728
[7]	LA GRASSA R, MARTELLATO E, CREMONESE G, et al. LU5M812TGT: an AI-powered global database of impact craters ≥ 0.4 km on the Moon[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2025, 220: 75-84 doi: 10.1016/j.isprsjprs.2024.11.010
[8]	赵丹冬. 月球着陆区小型撞击坑智能识别与空间分布特征分析[D]. 长春: 吉林大学, 2022 ZHAO D D. Intelligent Identification and Spatial Distribution Analysis of Small Craters in Lunar Landing Area[D]. Changchun: Jilin University, 2022
[9]	JIA M N, YUE Z Y, DI K C, et al. A catalogue of impact craters larger than 200 m and surface age analysis in the Chang’E-5 landing area[J]. Earth and Planetary Science Letters, 2020, 541: 116272 doi: 10.1016/j.jpgl.2020.116272
[10]	WANG Y X, NAN J, ZHAO C X, et al. A catalogue of impact craters and surface age analysis in the Chang’E-6 landing area[J]. Remote Sensing, 2024, 16(11): 2014 doi: 10.3390/rs16112014
[11]	XU W Y, CHEN Z L, ZHANG H F, et al. Automatic Martian polar ice cap extraction algorithm for remote sensing data and analysis of their spatiotemporal variation characteristics[J]. Remote Sensing, 2024, 16(7): 1201 doi: 10.3390/rs16071201
[12]	IQBAL W, HIESINGER H, BORISOV D, et al. Geological mapping and chronology of lunar landing sites: Apollo 14[J]. Icarus, 2023, 406: 115732 doi: 10.1016/j.icarus.2023.115732
[13]	IQBAL W, HIESINGER H, VAN DER BOGERT C H. Geological mapping and chronology of lunar landing sites: Apollo 12[J]. Icarus, 2020, 352: 113991 doi: 10.1016/j.icarus.2020.113991
[14]	CADOGAN P H. Automated precision counting of small lunar craters - a broader view[J]. Icarus, 2024, 408: 115796 doi: 10.1016/j.icarus.2023.115796
[15]	DELATTE D M, CRITES S T, GUTTENBERG N, et al. Automated crater detection algorithms from a machine learning perspective in the convolutional neural network era[J]. Advances in Space Research, 2019, 64(8): 1615-1628 doi: 10.1016/j.asr.2019.07.017
[16]	FAIRWEATHER J H, LAGAIN A, SERVIS K, et al. Automatic mapping of small lunar impact craters using LRO-NAC images[J]. Earth and Space Science, 2022, 9(7): e2021EA002177 doi: 10.1029/2021EA002177
[17]	HEAD III J W, FASSETT C I, KADISH S J, et al. Global distribution of large lunar craters: implications for resurfacing and impactor populations[J]. Science, 2010, 329(5998): 1504-1507 doi: 10.1126/science.1195050
[18]	CADOGAN P H. Automated precision counting of very small craters at lunar landing sites[J]. Icarus, 2020, 348: 113822 doi: 10.1016/j.icarus.2020.113822
[19]	YANG C, WANG X L, ZHAO D D, et al. Accurate mapping and evaluation of small impact craters within the lunar landing area[J]. Remote Sensing, 2024, 16(12): 2165 doi: 10.3390/rs16122165
[20]	KHANAM R, HUSSAIN M. YOLOv11: An overview of the Key Architectural Enhancements[OL]. arXiv preprint arXiv: 2410.17725, 2024
[21]	ZHUO S L, BAI H, JIANG L F, et al. SCL-YOLOv11: A Lightweight Object Detection Network for Low-illumination Environments[J]. IEEE Access, 2025, 13: 47653-47662 doi: 10.1109/ACCESS.2025.3550947
[22]	WANG G R, CHEN S Y, HU G, et al. Detection algorithm of abnormal flow state fluid on closed vibrating screen based on improved YOLOv5[J]. Engineering Applications of Artificial Intelligence, 2023, 123: 106272 doi: 10.1016/j.engappai.2023.106272
[23]	NAN J, WANG Y X, DI K C, et al. YOLOv8-LCNET: An Improved YOLOv8 Automatic Crater Detection Algorithm and Application in the Chang’E-6 Landing Area[J]. Sensors, 2025, 25(1): 243 doi: 10.3390/s25010243
[24]	AKYON F C, ALTINUC S O, TEMIZEL A. Slicing aided hyper inference and fine-tuning for small object detection[C]//Proceedings of the 2022 IEEE International Conference on Image Processing (ICIP). Bordeaux: IEEE, 2022: 966-970
[25]	ZHANG H, HAO C Y, SONG W R, et al. Adaptive slicing-aided hyper inference for small object detection in high-resolution remote sensing images[J]. Remote Sensing, 2023, 15(5): 1249 doi: 10.3390/rs15051249
[26]	JI J Z, GUO D J, LIU J Z, et al. The 1: 2, 500, 000-scale geologic map of the global moon[J]. Science Bulletin, 2022, 67(15): 1544-1548 doi: 10.1016/j.scib.2022.05.021
[27]	QIAN Y Q, XIAO L, ZHAO J W, et al. First magnetic and spectroscopic constraints on attenuated space weathering at the Chang’E-5 landing site[J]. Icarus, 2024, 410: 115892 doi: 10.1016/j.icarus.2023.115892
[28]	Meyer C. Lunar Sample Compendium[R]. NASA, 2005
[29]	李勃, 凌宗成, 张江, 等. 嫦娥三号着陆区月海玄武岩的年龄、成因及地质意义[J]. 岩石学报, 2016, 32(1): 19-28 LI Bo, LING Zongcheng, ZHANG Jiang, et al. Geochronology, petrogenesis and geological significance of the lunar basalts around CE-3 landing site[J]. Acta Petrologica Sinica, 2016, 32(1): 19-28
[30]	阳梅萍, 岳宗玉, 邸凯昌, 等. 基于全景相机数据的嫦娥四号着陆区次级坑统计分析[J]. 矿物岩石地球化学通报, 2021, 40(3): 720-729 YANG Meiping, YUE Zongyu, DI Kaichang, et al. Statistical analysis of secondary craters in the Chang’E-4 landing area based on panoramic camera data[J]. Bulletin of Mineralogy, Petrology and Geochemistry, 2021, 40(3): 720-729
[31]	LI Q L, ZHOU Q, LIU Y, et al. Two-billion-year-old volcanism on the moon from Chang’E -5 basalts[J]. Nature, 2021, 600(7887): 54-58 doi: 10.1038/s41586-021-04100-2
[32]	CUI Z X, YANG Q, ZHANG Y Q, et al. A sample of the Moon’s far side retrieved by Chang’E -6 contains 2.83-billion-year-old basalt[J]. Science, 2024, 386(6728): 1395-1399 doi: 10.1126/science.adt1093
[33]	LIU S C, DU K C, TONG X H, et al. In-situ mapping of iron and titanium with the Visible and Near-infrared Image Spectrometer (VNIS) along the Yutu-2 rover traverse on the far side of the moon[J]. Icarus, 2024, 412: 116003 doi: 10.1016/j.icarus.2024.116003
[34]	QIAO L, XU L Y, YANG Y Z, et al. Cratering records in the Chang’E -5 mare unit: filling the “age gap” of the lunar crater chronology and preparation for its recalibration[J]. Geophysical Research Letters, 2021, 48(22): e2021GL095132 doi: 10.1029/2021GL095132