Optimizing an image analysis protocol for ocean particles in focused shadowgraph imaging systems
A variety of imaging systems are in use in oceanographic surveys, and the opto-mechanical configurations have become highly sophisticated. However, much less consideration has been given to the accurate reconstruction of imaging data. To improve reconstruction of particles captured by Focused Shadow...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-07-01
|
| Series: | Frontiers in Marine Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmars.2025.1539828/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | A variety of imaging systems are in use in oceanographic surveys, and the opto-mechanical configurations have become highly sophisticated. However, much less consideration has been given to the accurate reconstruction of imaging data. To improve reconstruction of particles captured by Focused Shadowgraph Imaging (FoSI)—a system that excels at visualizing low-optical-density objects, we developed a novel object detection algorithm to process images with a resolution of ~ 12 μm per pixel. Suggested improvements to conventional edge-detection methods are relatively simple and time-efficient, and more accurately render the sizes and shapes of small particles ranging from 24 to 500 μm. In addition, we introduce a gradient of neutral density filters as a part of the protocol serving to calibrate recorded gray levels and thus determine the absolute values of detection thresholds. Set to intermediate detection threshold levels, particle numbers were highly correlated with beam attenuation (cp) measured independently. The utility of our method was underscored by its ability to remove imperfections (dirt, scratches and uneven illumination), and by capturing the transparent particle features such as found in gelatinous plankton, marine snow and a portion of the oceanic gel phase. |
|---|---|
| ISSN: | 2296-7745 |