Tools to Perform Local Dense 3D Reconstruction of Shallow Water Seabed

Abstract

Tasks such as distinguishing or identifying individual objects of interest require the production of dense local clouds at the scale of these individual objects of interest. Due to the physical and dynamic properties of an underwater environment, the usual dense matching algorithms must be rethought in order to be adaptive. These properties also imply that the scene must be observed at close range. Classic robotized acquisition systems are oversized for local studies in shallow water while the systematic acquisition of data is not guaranteed with divers. We address these two major issues through a multidisciplinary approach. To efficiently acquire on-demand stereoscopic pairs using simple logistics in small areas of shallow water, we devised an agile light-weight dedicated system which is easy to reproduce. To densely match two views in a reliable way, we devised a reconstruction algorithm that automatically accounts for the dynamics, variability and light absorption of the underwater environment. Field experiments in the Mediterranean Sea were used to assess the results.

Synchronization of Picam Cameras for Three-Dimensional Study of Dynamic Multi-Domains Natural Scenes

“…The matching algorithm used (Avanthey et al, 2016) relies on a self-adapting Harris point detector and on local statistical filtering on the vector flow formed by the matched points to discard outliers (bad matches). The results of our tests show that we obtain an average rate of 45% of good matches (inliers) on all of our synchronized pairs.…”

“…Among them, the GoPros, uEye and PiCam v2 sensors fit particularly well our constraints. The PiCam v2 of Raspberry Pi which was released in 2016 (3280 × 2464px, 1.4µm pixel size, ∼30e) seems to offer a good compromise between the two other sensors tested by (Avanthey et al, 2016) for dynamic environments: the uEye camera from IDS (entry-level professional sensor, 1280 × 1024px, 5.3µm pixel size, ∼500e) and the Hero 2 camera from GoPro (a sports camera in the consumer category, 3840×2880px, 1.6µm pixel size, ∼200e, similar today to a GoPro Hero 7 Silver Edition). In terms of cost and weight, the PiCam is far below these two cameras, even by adding a board for control and data storage (Raspberry Pi Zero or 3+ for example: 10 to 40e).…”

“…Studies carried out by (Venkataraman et al, 2013, Santise et al, 2017, Piras et al, 2017 shows that the PiCam sensor is suitable for photogrammetric work. Simultaneous acquisitions (mechanical or, better, electronic trigger signal), which lead to synchronizations at best at 500ms, are rarely sufficient with regard to the dynamism of natural areas as we can see in (Avanthey et al, 2016) with the GoPro. In terms of control, the PiCam gets closer to the uEye camera and lets hope for a synchronization time as good, or even better, than the one obtained for the latter (5ms) by (Beaudoin et al, 2015) without its software instability problems related to the driver.…”

“…To reconstruct dynamic objects (fauna or flora for example), we need to freeze the movement. In other words, the local relative displacement taking place between two shots should not exceed the size of the GSD (Avanthey et al, 2016). This is possible if the following two conditions are fulfilled: several sensors are required and they must be synchronized.…”

Surveying the Underwater Arcaheological Site of Cape Glaros At Pagasetikos Gulf

“…Such an aspect of visualized data can serve most archaeological research needs, from analysis of the detailed overview of the site, to impressive presentations. Such developed work has shown its functionality and effectiveness so far not only in cases of underwater archaeological projects, but also in research projects of digital cultural heritage in general (Drap 2001, Allen 2004, Avanthey 2016.…”