Abstract:
The objective of this paper is to describe and compare the use of different high-resolution survey techniques to create and analyse 3D surface models made up of point clouds, using open-source and freeware post-processing tools. Within this framework, several tests were conducted on a three-dimensional scale model of a breakwater, with surveys performed by a smartphone, the Microsoft® Kinect® sensor and a UAV camera. All the tests were conducted for a physical model placed in an empty wave tank. For this comparison, a number of post-processing software were used: the Microsoft Kinect Fusion Application, Visual SFM and MicMac, all freeware software. The clouds of points were compared using the CloudCompare freeware software and the Golden Software Surfer commercial software.
In order to obtain a clear comparison between these tools (i.e., to better evaluate accuracy and associated errors), all the point clouds (with RGB colour information) were aligned with data coming from a survey used to coordinate some points in the model (Ground Control Points). All the clouds were compared with a corrected cloud based upon coordinated points painted on the concrete floor of the wave tank.
The specific experimental conditions needed for using each of these methodologies were investigated and classified.