An innovative semantically guided SAR imaging and target enhancement method
Abstract Conventional sparse synthetic aperture radar (SAR) imaging methods apply regularisation to constrain scene priors. However, these methods often neglect specific target regions, resulting in undifferentiated imaging. This letter introduces a novel network for sparse SAR imaging and target re...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-12-01
|
| Series: | Electronics Letters |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/ell2.70123 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Conventional sparse synthetic aperture radar (SAR) imaging methods apply regularisation to constrain scene priors. However, these methods often neglect specific target regions, resulting in undifferentiated imaging. This letter introduces a novel network for sparse SAR imaging and target region enhancement, using target segmentation semantic information to integrate a modified attention mechanism into the image formation process. The complex‐valued convolution is used to handle SAR's complex‐valued data. Additionally, SAR imaging operator and its inverse operator accelerate echo processing. Experiments show improvements in the target–background ratio (TBR) and overall reconstruction quality across various downsampling scenarios. |
|---|---|
| ISSN: | 0013-5194 1350-911X |