Dataset

Machine Learning for Synthetic Aperture Radar Autofocus


Dataset

Please see AutofocusSAR github or AutofocusSAR webpage for more details.

ALOS PALSAR

Obtain RAW SAR data

Nine RAW SAR images (L1.0) are acquired from https://asf.alaska.edu/. The data can be easily obtained by searching the scene name (ALPSRP020160970, ALPSRP054200670, ALPSRP103336310, ALPSRP110940620, ALPSRP115120970, ALPSRP268560540, ALPSRP269950430, ALPSRP273680670, ALPSRP278552780) in https://search.asf.alaska.edu/. The worldmaps of these images are illustrated as follows:

     
Vancouver, ALPSRP020160970 Vancouver, ALPSRP020160970 Xian, ALPSRP054200670 Xian, ALPSRP054200670 Hawarden, ALPSRP103336310 Hawarden, ALPSRP103336310
Hefei, ALPSRP110940620 Hefei, ALPSRP110940620 Langley, ALPSRP115120970 Langley, ALPSRP115120970 Florida, ALPSRP268560540 Florida, ALPSRP268560540
Kaliganj, ALPSRP269950430 Kaliganj, ALPSRP269950430 SimiValley, ALPSRP273680670 SimiValley, ALPSRP273680670 Toledo, ALPSRP278552780 Toledo, ALPSRP278552780

Making Dataset

The download nine sar raw images are imaged by Range Doppler Algorithm (RDA). Since the orignal image is very large, sub-region with size of $8192×8192$ is selected for each image. The selected regions are shown as follows:

     
Vancouver, ALPSRP020160970 Vancouver, ALPSRP020160970 Xian, ALPSRP054200670 Xian, ALPSRP054200670 Hawarden, ALPSRP103336310 Hawarden, ALPSRP103336310
Hefei, ALPSRP110940620 Hefei, ALPSRP110940620 Langley, ALPSRP115120970 Langley, ALPSRP115120970 Florida, ALPSRP268560540 Florida, ALPSRP268560540
Kaliganj, ALPSRP269950430 Kaliganj, ALPSRP269950430 SimiValley, ALPSRP273680670 SimiValley, ALPSRP273680670 Toledo, ALPSRP278552780 Toledo, ALPSRP278552780

Table3: Information of selected regions.

Number Scence name Area Lefttop pixel index Effective velocity(m/s) PRF(Hz)
1 ALPSRP020160970 Vancouver (10000, 3600) 7153 1912.0459
2 ALPSRP054200670 Xi’an (16000, 1000) 7185 2159.8272
3 ALPSRP103336310 Hawarden (10000, 1000) 7211 2105.2632
4 ALPSRP110940620 Hefei (18000, 1000) 7188 2145.9227
5 ALPSRP115120970 Langley (10000, 2100) 7174 2155.1724
6 ALPSRP268560540 Florida (3000, 2000) 7190 2159.8272
7 ALPSRP269950430 Kaliganj (18000, 1000) 7195 2159.8272
8 ALPSRP273680670 SimiValley (15000, 2000) 7185 2155.1724
9 ALPSRP278552780 Toledo (15000, 1000) 7178 2141.3276

To make defocused data, different effective velocities are applied to compute the matched filter in azimuth when executing the range-Doppler algorithm. Image patches with size 256×256 are selected from the above sub-regions to create the final dataset. Please see our paper for details. The final dataset can be downloaded from MEGA part1 (part1), MEGA part2 or BaiduYunPan all parts (accessed on 13 August 2021), the extraction code is d7fk.

Citation

If you find the dataset is useful, please kindly cite our paper and star our pakcage AutofocusSAR on GitHub:

@article{Liu2021Fast,
  title={Fast SAR Autofocus Based on Ensemble Convolutional Extreme Learning Machine},
  author={Liu, Zhi and Yang, Shuyuan and Feng, Zhixi and Gao, Quanwei and Wang, Min},
  journal={Remote Sensing},
  volume={13},
  number={14},
  pages={2683},
  year={2021},
  publisher={Multidisciplinary Digital Publishing Institute},
  doi={https://doi.org/10.3390/rs13142683}
}