# DeepCraters Lunar Craters Database 2020-07 This is a database of 117240 lunar impact craters larger than about 1 km in diameter. Craters were identified on 120m CE-1, 50m CE-2 DOM and DEM images captured by Chang’E-1 (CE-1) and Chang’E-2 (CE-2) orbiters. And that the formation Systems of new detected craters, 18,996 larger than 8km, are estimated.
## Method review The database of 117240 lunar impact craters are identified by the trained R-FCN model and the age of craters is estimated by a dual-channel classification model with the semisupervised method. (the details of method can be found in our paper[1]).
## Briefly Description The database contains two comma-delimited table files: `Lunar_Crater_Database_DeepCraters_2020.csv` and `Aged_Lunar_Crater_Database_DeepCraters_2020.csv`.
> The first table is the detected craters larger than about 1 km in diameter, which contains 117240 craters with 5 common fields in each crater. What needs illustration is that the first 7009 craters contains additional 2 fields `Lat_new` and `Lon_new` because of the inaccuracy of the CE1 dataset. the details of each field are listed below: > - `Flags_data`: 'CE1' or 'CE2' indicats the crater detected from CE-1 or CE-2 data; > - `ID`: the index of craters; > - `Lat`: the latitude of crater center in degree; > - `Lon`: the longitude of crater center in degree; > - `Diam_km`: the diameter of crater in kilometer; > - `Lat_new`: the latitude of the corresponding crater in Robbins Lunar Crater Database[2] in degree; > - `Lon_new`: the longitude of the corresponding crater in Robbins Lunar Crater Database[2] in degree. >
> The second table is the aged craters larger than 8 km in diameter, which contains 18996 craters with 6 fields in each craetr. And the first 5 fields are the same as in the first table, the last field `Age` indicates the system that the crater belongs to. > > - `Age`=1: pre-Nectarian System > - `Age`=2: Nectarian System > - `Age`=3: Imbrian System > - `Age`=4: Eratosthenian System > - `Age`=5: Copernican System
--------
## Note
The diameter of craters detected by DeepCraters is slightly larger than the real size of the craters.
## Reference [1] Yang C , Zhao H , Bruzzone L , et al. Lunar impact craters identification and age estimation with Chang'E data by deep and transfer learning. [Online]. https://nature-research-under-consideration.nature.com/posts/57165-lunar-impact-craters-identification-and-age-estimation-with-chang-e-data-by-deep-and-transfer-learning.
[2] 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. Journal of Geophysical Research (Planets),124(4), 871-892(2019).
Funding
This study was supported by the CE-1 and CE-2 mission of CLEP. It was also supported by the National Natural Science Foundation of China grant 61572228 and 61972174, the Science-Technology Development Plan Project of Jilin Province of China grant 20190303006SF and 20190302107GX and the Industrial Innovation Special Funds Project of Jilin Province grant 2019C053-5 and 2019C053-7.