Individual underwater images taken close to the seabed (~1-2m) have high resolution and minimal water column attenuation, but cover only a small area. A landscape mosaic is a composite of many underwater images. The mosaics have the clarity and resolution of individual pictures but afford a "landscape view" of the seabed (Fig 1). The U.S. Strategic Environmental Research and Development Program (SERDP) has supported a) the development of software tools for generating underwater landscape mosaics without relying on external navigation and b) the evaluation of these mosaics for coral reef mapping and monitoring. We are seeking to identify potential applications and partners who might take advantage of this unique new data source. Data Acquisition Requirements: Mosaics are made in one of two modes: "Standard mode" uses video data only; "Enhanced mode" uses still images acquired synchronously with the video. Both need: Near-nadir view video 1-2 m from seabed. High (~80 percent) overlap between swaths. Enhanced mode additionally requires: Mosaic Characteristics:
Figure 1: Mosaic overview: Video images acquired by a diver (A) or other platform such as an ROV (B) are automatically stitched together to form a landscape mosaic (C) covering a large area (about 100 m2 in this case). "Standard mode" (i.e. video only) produces mosaics with mm-scale resolution (D). In "enhanced mode", still imagery is acquired simultaneously with the video (E) to achieve sub-mm resolution (F).
Still camera synchronized with video.
Area covered: ~ 400 m^2 (~2000 frames) Spatial resolution (pixel size): sub-mm (in enhanced mode); ~ 3 mm (in standard mode). Spatial accuracy: +/-5 cm (1 standard deviation)
Data acquisition takes about 20-40 minutes (i.e. one dive), depending on the size of the site to be surveyed. The mosaic production is highly automated, requiring about 4 man-hours and 24-36 hours computer time with current desktop processors.
Mosaics retain key strengths of a diver-based approach, while overcoming the limitations of diver-based or photo-quadrat / video transect methods (Table 1). Table 1: Comparison of Monitoring Techniques Green indicates full capability, yellow partial capability, and red poor capability. Note (1): Enhanced mode required for species-level IDs, but identification of major functional groups (e.g., corals, sponges, algae) is done with standard mode. Note (2): Enhanced mode required.
Mosaics are made in one of two modes: "Standard mode" uses video data only; "Enhanced mode" uses still images acquired synchronously with the video. Both need:
Enhanced mode additionally requires:
Data acquisition takes about 20-40 minutes (i.e. one dive), depending on the size of the site to be surveyed. The mosaic production is highly automated, requiring about 4 man-hours and 24-36 hours computer time with current desktop processors.
Mosaics retain key strengths of a diver-based approach, while overcoming the limitations of diver-based or photo-quadrat / video transect methods (Table 1). Table 1: Comparison of Monitoring Techniques
Green indicates full capability, yellow partial capability, and red poor capability. Note (1): Enhanced mode required for species-level IDs, but identification of major functional groups (e.g., corals, sponges, algae) is done with standard mode. Note (2): Enhanced mode required.
Figure 2: Mosaic of a scar created by a ship grounding on a shallow reef, Florida Keys (depth = 3 m). The dashed line marks the extent of damage. The inset shows this mosaic inserted into Google Earth, illustrating the potential to incorporate mosaics in GIS systems. Groundings are large and cumbersome to survey solely by divers. An image conveys more information about the extent of the damage than measurements of the overall dimensions, especially when viewed by non-technical personnel (e.g. juries).