Photo mosaic with different taxa of benthic filter feeders growing on a vertical wall at 430 m depth in the fjord of Trondheim, Norway. Living stony corals appear white. Dead stony corals appear grey due to loss of antibacterial mucus production and consequent silt sedimentation. The area covered by the photo mosaic in the left image is approximately 2 m x 5 m. 

Contexts in source publication

Context 1

... make the mosaic appear seamless, it is necessary to avoid steps in image intensity on the borders between the images. After the topology has been estimated, intensity steps are reduced by merging the images in a process called blend- ing (based on Burt and Adelson, 1983). In the first step of the blending process, the original images are decomposed in the frequency domain using band-pass filters to form several frequency component images. These images are then merged over transition zones within the overlay areas of the neighboring images. The center of this transition zone is a line in the middle between the centers of the images. The resulting pixel intensity in each frequency-component image in the transition zone is determined by applying a weighted average of the overlapping images. To remove the bor- der between two images, the average is weighted by the distance to the center line of the transition zone. A mosaic is built for each frequency component before they are joined in the final step of the blending procedure. By choosing a narrow transition zone for the high-frequency components, and a wider transition zone for the lower- frequency components, double appear- ance of smaller objects is avoided, while the larger structures in the initial image set are merged smoothly; the resulting mosaic thus appears seamless. The area shown in Figure 3 is located at 430-m depth and is approximately 2-m wide and 5-m high. The tempera- ture was 8.1°C and salinity was 35 psu at the site. It is exposed to the Norwegian coastal current (Sakshaug and Sneli, 2000), and high abundance of zoo- plankton was observed at the site. The photo mosaic shows a benthic filter- feeder community comprised of benthic cold-water corals, bivalves, sponges, and numerous other taxa. Abundant horny corals like the Paragorgia arborea and Paramuricia placomus are attached to the exposed rock wall. The dominant species are the stony coral Lophelia pertusa and the European giant file clam Acesta exca- vata . The bivalve appears in large numbers, found in high density on the rock wall beneath the hanging stony coral branches. Both living and dead stony corals are visible. Mucus containing anti- biotic compounds, which also removes sediment and particles (Mortensen, 2000), makes living stony corals look “clean and white.” Dead stony coral, in which mucus production has ceased, is grey and brown due to sediment accu- mulation, microfauna attachment, and invasion by boring sponges. The horizontal extent of the area shown in Figure 4 is approximately 18 m and the vertical extent is about 8 m. The depth is about 390 m. At this site, horny corals, stony corals, bivalves, rockfish, and actinia are among the most detect- able taxa. Large branches of stony corals are attached to the vertical wall, and again both dead and living corals are dis- tinguishable. The bivalves reside in clus- ters both on exposed rock and at sites protected by stony coral branches. The taxonomic diversity on the vertical rock wall appears to be remarkably high compared to other habitats at the same depth in the study region. Due to the near-vertical inclination, sediment does not accumulate on the rock wall, which can be advantageous for many filter feeders. A reduced risk of being buried by sediment or marine snow may be favorable for larval settlement and for slowly growing organisms. For the larger mobile biota, attached taxa, such as corals, represent shelter. The photo mosaic in Figure 5 shows a shipwreck at approximately 170-m depth. The area depicted was approximately 37-m long and 19-m wide. The wreck is eroded and most of the wood is decomposed, but the lower part of the wooden hull remains. The photo mosaic was produced for post-investigation documentation of the site before the investigation was closed. Two cannons, four lead cable holes, a large number of bottles, and the wooden remnants of the hull are the most dis- tinct objects in the photo mosaic. The keel line is visible and indicates that the bow was probably located close to the leaden cable holes, while the stern was near the bottles. From observation of the wreck site and the artifacts recov- ered, archaeologists estimate that the wreck is most likely a merchant ship that was involved in trade between Europe and Russia. It probably sank in the early 1800s (Søreide and Jasinski, 2005). The Figure 6 photo mosaic illustrates the wreck site after 0–20 cm of sediment was removed by a suction unit mounted on the WROV. Inset (a) of Figure 6 shows a three-dimensional processed model of a mug in the site to elucidate how the geometry of objects can be reconstructed by photogrammetry. Insets (b), (c), and (d) illustrate how the photo mosaic can be magnified to reveal details of the site. The photo mosaic shows exposed wooden hull parts, bottles, bottle necks sticking out of the sand, dishes, mugs, and other artifacts from the ship. Bottle shapes include round and square; there are bottles made of glass, and bottles made of ceramic mate- rial. In the middle of the area, there are wooden planks remaining from the bottom and keel of the ...

Context 2

... the lights were mounted on booms to increase the lateral distance between light source and camera, with the aim of reducing backscatter light to the camera (see Figure 1). The lateral distance between the camera and light source was approximately 0.8 m. On the WROV, a lateral separation of light sources and camera of 0.7 m was obtained without light booms. During the marine biological survey, the camera was mounted on the underwater vehicle with a horizontal image axis (i.e., aiming orthogonally at the rock wall). For the marine archaeological survey, the camera pointed downward, again imaging its target, the seabed, orthogonally. The camera used in both projects was a 12-bit dynamic range camera (Uniqvision) that captures one image every four seconds. Optimal image overlap, sidelap, seabed resolution, and acquisition efficiency are essential to mosaic quality. They are achieved by carefully choosing proper altitude, line spacing, and velocity when planning the survey. The term “over- lap” is used for the common area in two images taken sequentially while the camera moves along a line, and “sidelap” denotes the common area of two images across track. The area covered by each image directly depends on the altitude of the vehicle and the camera’s field- of-view angle. A general recommenda- tion for mosaics of aerial photography is to use 50% overlap and 25% sidelap. Practice has shown that this trade-off between redundant image points in image pairs and the need for effective data acquisition is also reasonable for underwater photo mosaics. A 45° field of view was used in all data-acquisition operations described in this paper. This narrow field of view produces low obliqueness in images. Apart from the available propulsion power on the underwater vehicle, the survey speed is limited by the overlap constraint, covered area, and how often it is possible to take images during data acquisition. The potential for motion blur in the images captured was not considered when the theoretical speed was calculated. During data acquisition for the marine biological study, the ROV followed horizontal paths along a vertical wall, with each path one meter shallower than the previous one, so that the rock wall was covered in a back-and-forth pattern similar to mowing a lawn. The ROV heading was kept constant so that the cameras always pointed toward the wall while the vehicle moved sideways. The distance from the camera to the seafloor or wall target was approximately 2 m. This resulted in a coverage area of 2.3 m 2 for each image frame. The camera produced 1024 x 1024-pixel images, with each pixel covering 2.2 mm 2 of rock wall. This resolution enables identifica- tion of specimens a few centimeters long, depending on the distinctness of their morphological characteristics. In order to obtain a 50% overlap, we calculated that the velocity would theo- retically be limited to 0.2 m s -1 . Heading deviations resulted in overlap variations, and to compensate for this, the speed was lowered to 0.15 m s -1 . The vertical line spacing was one meter, providing a sidelap of 34%. The theoretical data collection efficiency was 800 m 2 h -1 for this arrangement. The photo mosaic in Figure 3 is made from nine images, while the photo mosaic in Figure 4 is constructed from 73 images. To cover the complete wreck site, the WROV was maneuvered along eleven 50-m-long survey lines in a lawn-mower pattern by the closed-loop control system. The camera altitude was approximately 1.9 m to ensure sufficient lighting and area coverage of 1.6 x 1.6 m. A line spacing of 1 m resulted in 36% sidelap. Approximately 550 images were used in the resulting photo mosaic shown in Figure 5. Due to varying current conditions at the site, it was difficult to keep the vehicle velocity constant and hence the velocity was kept lower than the theo- retical 0.36 m s -1 to ensure overlap. In the data set collected with the WROV installed in the excavation support frame (Figure 2), the vehicle was moved in a in a parallel or back-and- forth, lawn-mower pattern with 0.5 m line spacing. The distance between each image along track was also 0.5 m. The altitude varied from 1.5–1.7 m due to an excavated pit in the middle of the area. The resulting overlap and sidelap were close to 50%. The large sidelap enabled photogrammetry processing of the images. A total of 270 images were included in the photo mosaic in Figure ...