Image Restoration of Highly Reflective Polarization Targets in Turbid Media Based on Circularly Polarized Light
The traditional underwater active polarization imaging system primarily employs linearly polarized light as its illumination source. However, in cases where the target exhibits high light polarization upon reflection in a turbid medium, the difference between the polarization of the targetȁ...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10737088/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The traditional underwater active polarization imaging system primarily employs linearly polarized light as its illumination source. However, in cases where the target exhibits high light polarization upon reflection in a turbid medium, the difference between the polarization of the target’s reflected light and that of the backscattered light in the turbid water becomes small. This can result in the backscattered light intensity sampled per unit pixel exceeding that of the target, significantly degrading image quality. To address this issue, this paper introduces an underwater target imaging method based on active circularly polarized light. This method capitalizes on the superior polarization-preserving characteristics of circularly polarized light relative to linearly polarized light, along with the spin reversal of circularly polarized reflected light. Based on these properties, combined with circularly polarized differential imaging method and image frequency domain filtering method, the method in this paper can realize the recovery of target images with highly reflected polarized light in turbid waters. The performance of the algorithm described in this paper is assessed through simulation experiments and validated via laboratory tests, demonstrating the superior effectiveness of the proposed method compared to previous approaches. |
|---|---|
| ISSN: | 2169-3536 |