Preliminary assessment of Photogrammetric Approach for detailed dimensional and colorimetric reconstruction of Corals in underwater environment
2018
The use of 3D scan technologies to provide 3D color models of some bio-constructor organisms without removing them from the sea represents a valuable contribution for the preservation of current marine flora and fauna. The necessity to perform accurate measurements of volume, surface area and other morphometric characteristics is to increase the knowledge about the habitat-forming species. In this way, it is possible to describe the modification they made to the seabed topography, to monitor their shape modification according to local environmental conditions and evaluate 3D structure and color of some species sensitive to climate changes. The necessity of working in an environment characterized by different light conditions and particulate concentrations, the need of having high-performance sealing and, last but not least, the portability guarantee have limited the range of possible measurement techniques among those available for 3D shape reconstruction. Within this context, the idea underlying this research activity is to adopt photogrammetric technique for underwater surveys. In particular, a comprehensive and detailed analysis of a small object in and outside-water is performed in lab-scale tests by means of two different camera lenses, to pinpoint the effectiveness and the weakness of the solution identified. Another contribution of this paper is the uses of a color correction algorithm in order to correct the scene illumination, improving the 3D reconstruction of the detected objects. It is shown the capability of the algorithm to solve the color mismatch problem in multi-view color images. Results show the 3D models with details very well reconstructed both in and outside-water. The image pre-processing step drastically improve the matching points recognition and the reconstruction of the whole object independently to environmental illumination changes.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
11
References
1
Citations
NaN
KQI