Fig 1.
A) Geolocation of Faafu Atoll in the Maldives. B) Geolocation of Magoodhoo island in the Faafu Atoll. C) Diving sites and coral reefs in the vicinity of Magoodho island. Sampled sites: 1. Wall Street 2. Magoodhoo Home Reef, left side 3. Dharanboodhoo Reef Bis 4. Dhigu Reef 5. Maaga.
Fig 2.
Massive coral growth forms sampled in this study (specie, location, date sampled).
A) Gardineroseris planulata, Wall Street, 08/05/14 B) Porites lutea, Home Reef, 08/05/14 C) 8. Platygyra daedalea, Dhigu, 13/05/14 D) Porites lobata, Darambodhoo, 12/05/14. Scale bar 10 cm. Note: The coloured chart in some of the images is the ColorChecker Passport used for colour calibration and scale reference.
Fig 3.
Branched and tabular coral growth forms sampled in this study (specie, location, date sampled).
A) Acropora sp. 1, Home Reef, 08/05/14 B) Pocillopora verrucosa, Maaga, 17/05/14 C) Acropora sp. 2 (cf valida), Darambodhoo, 12/05/14 D) Acropora hyacinthus (tabular) Darambodhoo, 12/05/14. Scale bar 10 cm. Note: The coloured chart in some of the images is the ColorChecker Passport used for colour calibration and scale reference.
Fig 4.
Encrusting and foliose coral growth forms sampled in this study (specie, location, date sampled).
A) Diploastrea heliopora (encrusting), Home Reef, 08/05/14 B) Echinopora lamellosa (foliose), Wall Street, 17/05/14. Scale bar 10 cm.
Fig 5.
Dotted boxes represent the steps of this methodology, rounded boxes represent digital actions, ovals represent physical actions, line arrows represent the flow of actions and dotted arrows represent reiterations (repeating underwater imaging after sampling and repeating dry imaging after bleaching skeletons).
Table 1.
Detail of criteria used for digital analysis of Surface Area and Volume of 3D models.
Fig 6.
Digitisation pipeline of 3D coral sets.
A) Unprocessed digital photograph of underwater Gardineroseris planulata colony extracted from video and used in model generation. B) 3D model at community scale of same underwater set where G. planulata colony was segmented from. C) 3D model at colony scale showing extraction of sample from G. planulata colony. D) 3D model at subcolony level of G. planulata using dry imaging on sample with live tissue displaying section where detailed macro imaging was performed. E) Detail of 3D macro imaging of live tissue sample of G. planulata. F) 3D model at subcolony level of G. planulata using dry imaging on same sample of bleached skeleton displaying section where detailed macro imaging was performed. G) Detail of 3D macro imaging of bleached skeleton sample of G. planulata.
Table 2.
Digital specifications in coral imaging.
Fig 7.
Multi-image photogrammetry of underwater coral colonies and segmentation.
A) 3D model of community around Gardineroseris planulata colony. B) Set of camera points after alignment. C) Segmentation of colony for Unsampled calculations. D) Segmentation of colony for TSA and TV Sampled calculations. E) Segmentation of colony for ESA “Sampled” calculations. Dashed line represents area or outline of segmented portion of 3D model.
Fig 8.
Dry multi-image photogrammetry of coral samples and segmentation.
A) Set of camera points after alignment of Gardineroseris planulata imaging sample. B) PC = 3D point cloud. TM = Triangle mesh. SM = Untextured surface mesh. TS = Textured and refined surface mesh. C) Segmentation of 3D sample for TSA and TV Tissue and Skeleton calculations. D) Segmentation of 3D sample for ESA Tissue and Skeleton calculations. Dashed line represents outline of segmented portion of 3D model.
Fig 9.
Macro detail of 3D models for taxonomical imaging.
A) Textured 3D surface mesh of Gardineroseris planulata. B) Same untextured 3D surface mesh of Gardineroseris planulata. Scale bar represents 7.6 mm. C) Textured 3D surface mesh of Echinopora lamellosa. D) Same untextured 3D surface mesh of Echinopora lamellosa. Scale bar represents 6.8 mm. E) Textured 3D surface mesh of Pocillopora verrucosa. F) Same untextured 3D surface mesh of Pocillopora verrucosa. Scale bar represents 4.4 mm. G) Textured 3D surface mesh of Porites lutea. H) Same untextured 3D surface mesh of Porites lutea. Scale bar represents 3.6 mm.
Table 3.
Surface area measurements from 3D models of corals generated by underwater and dry multi-image photogrammetry.
Table 4.
Volume measurements from 3D models of corals generated by underwater and dry multi-image photogrammetry.
Table 5.
Approximate time used to generate a 3D model and its area surface and volume calculations, and water volume estimation of bleached skeletons samples of each colony.
Fig 10.
Over and underestimations in 3D models.
A) Acropora sp. 1. B) Pocillopora verrucosa. C) Platygyra daedalea. D) Acropora hyacinthus. 1) Photographs from imaging session. 2) Texture mapped 3D model with shading of fresh sample. 3) Untextured 3D model of fresh sample. 4) Untextured 3D model of bleached skeleton.
Fig 11.
3D Models of fresh coral samples and bleached skeletons.
A) Gardineroseris planulata. B) Porites lutea. C) Platygyra daedalea. D) Pocillopora verrucosa. 1) 3D Models with biomass. 2) Bleached skeleton 3D models. Black arrow represents location of sponge epiphyte before bleaching.
Fig 12.
A) Acropora hyacinthus. B) Echinopora lamellosa. C) Pocillopora verrucosa. D) Porites lutea. 1) 3D environment with texture map around target colony. 2) 3D model of segmented colony. 3) 3D model of representative sample from colony. 4) Corallite detail in 3D model using close-range photogrammetry.