Enhanced Underwater Image Restoration and Clarity Optimization Using Advanced Digital Image Processing Techniques
Underwater image enhancement is essential in various tasks such as underwater detection, surveillance, and exploration given that the images these activities produce usually have poor quality because of the influence of some factors (low contrast, noise, and color distortion caused by light absorpti...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2025-01-01
|
| Series: | EPJ Web of Conferences |
| Online Access: | https://www.epj-conferences.org/articles/epjconf/pdf/2025/13/epjconf_icetsf2025_01027.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Underwater image enhancement is essential in various tasks such as underwater detection, surveillance, and exploration given that the images these activities produce usually have poor quality because of the influence of some factors (low contrast, noise, and color distortion caused by light absorption and scattering). This paper suggests a new way of Digital Image Processing (DIP) technique that combines Dark Channel Prior (DCP), Gray-World White Balancing (GWWB), and Contrast-Limited Adaptive Histogram Equalization (CLAHE) to realize the functions of dehazing, color correction, and contrast enhancement effectively. Furthermore, Gaussian bilateral filtering and Rayleigh stretching are utilized as denoising methods and edge clarity improvement. From the outcomes of the experiment based on PSNR and SSIM metrics, it is evident that the method presented in this paper has better performance than the traditional algorithms. Also, the presented method if free from training samples, thus allowing the benefits of computationally efficient and real-time processing that machine learning models wouldn't have. For this reason, the proposed method is particularly useful in the following areas: marine life observation, underwater robotics, and archaeological reconnaissance. In general, this approach gives low-cost, versatile, high-quality underwater image enhancement. |
|---|---|
| ISSN: | 2100-014X |