@article{Agrafiotis2019a,

title = {DepthLearn: Learning to correct the refraction on point clouds derived from aerial imagery for accurate dense shallow water bathymetry based on SVMs-fusion with LiDAR point clouds},

author = {Panagiotis Agrafiotis and Dimitrios Skarlatos and Andreas Georgopoulos and Konstantinos Karantzalos},

url = {https://www.mdpi.com/2072-4292/11/19/2225},

doi = {10.3390/rs11192225},

issn = {20724292},

year  = {2019},

date = {2019-09-01},

journal = {Remote Sens.},

volume = {11},

number = {19},

pages = {2225},

publisher = {MDPI AG},

abstract = {The determination of accurate bathymetric information is a key element for near offshore activities; hydrological studies, such as coastal engineering applications, sedimentary processes, hydrographic surveying, archaeological mapping and biological research. Through structure from motion (SfM) and multi-view-stereo (MVS) techniques, aerial imagery can provide a low-cost alternative compared to bathymetric LiDAR (Light Detection and Ranging) surveys, as it offers additional important visual information and higher spatial resolution. Nevertheless, water refraction poses significant challenges on depth determination. Till now, this problem has been addressed through customized image-based refraction correction algorithms or by modifying the collinearity equation. In this article, in order to overcome the water refraction errors in a massive and accurate way, we employ machine learning tools, which are able to learn the systematic underestimation of the estimated depths. In particular, an SVR (support vector regression) model was developed, based on known depth observations from bathymetric LiDAR surveys, which is able to accurately recover bathymetry from point clouds derived from SfM-MVS procedures. Experimental results and validation were based on datasets derived from different test-sites, and demonstrated the high potential of our approach. Moreover, we exploited the fusion of LiDAR and image-based point clouds towards addressing challenges of both modalities in problematic areas.},

keywords = {Aerial imagery, Bathymetry, Data integration, Fusion, LiDAR, Machine Learning, Point Cloud, Refraction effect, Seabed Mapping, SVM, UAV},

pubstate = {published},

tppubtype = {article}

}

@article{Agrafiotis2019,

title = {SHALLOW WATER BATHYMETRY MAPPING from UAV IMAGERY BASED on MACHINE LEARNING},

author = {P Agrafiotis and D Skarlatos and A Georgopoulos and K Karantzalos},

url = {https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLII-2-W10/9/2019/},

doi = {10.5194/isprs-archives-XLII-2-W10-9-2019},

issn = {21949050},

year  = {2019},

date = {2019-04-01},

journal = {ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci.},

volume = {42},

number = {2/W10},

pages = {9--16},

abstract = {The determination of accurate bathymetric information is a key element for near offshore activities, hydrological studies such as coastal engineering applications, sedimentary processes, hydrographic surveying as well as archaeological mapping and biological research. UAV imagery processed with Structure from Motion (SfM) and Multi View Stereo (MVS) techniques can provide a low-cost alternative to established shallow seabed mapping techniques offering as well the important visual information. Nevertheless, water refraction poses significant challenges on depth determination. Till now, this problem has been addressed through customized image-based refraction correction algorithms or by modifying the collinearity equation. In this paper, in order to overcome the water refraction errors, we employ machine learning tools that are able to learn the systematic underestimation of the estimated depths. In the proposed approach, based on known depth observations from bathymetric LiDAR surveys, an SVR model was developed able to estimate more accurately the real depths of point clouds derived from SfM-MVS procedures. Experimental results over two test sites along with the performed quantitative validation indicated the high potential of the developed approach.},

keywords = {Bathymetry, Machine Learning, Point Cloud, Refraction effect, Seabed Mapping, SVM, UAV},

pubstate = {published},

tppubtype = {article}

}

@article{Maltezos2018,

title = {Plane detection of polyhedral cultural heritage monuments: The case of tower of winds in Athens},

author = {Evangelos Maltezos and Charalabos Ioannidis},

url = {https://linkinghub.elsevier.com/retrieve/pii/S2352409X17305126},

doi = {10.1016/j.jasrep.2018.03.036},

issn = {2352409X},

year  = {2018},

date = {2018-06-01},

journal = {J. Archaeol. Sci. Reports},

volume = {19},

pages = {562--574},

publisher = {Elsevier Ltd},

abstract = {This study introduces an efficient and easy to implement plane detection approach towards the extraction of high-level information from 3D point clouds associated with polyhedral cultural heritage monuments. An adapted version of the randomized Hough transform (RHT) called “adaptive point randomized Hough transform” (APRHT) and a multiscale framework in terms of Level of Detail 1 (LoD 1) and LoD 2 are proposed. A dense image matching point cloud of an octagonal tower called Tower of Winds, which is situated on the northern foot of the Acropolis hill in Athens was used. A pre-process is carried out to extract points associated with the vertical structural elements. Then a plane detection process is performed in terms of LoD 1 to calculate the plane parameters ($theta$, $phi$ and $rho$) of each of the eight planar surfaces using a coarse form of the entire monument, that is, a sparse point cloud extracted via subsampling process. A mask of one representative detected planar surface is used to clip the initial point cloud with the initial point density. Then, a second plane detection process in terms of LoD 2 at the clipped point cloud is implemented to calculate the corresponding accurate plane parameters. The results are useful for cultural heritage preservation purposes and illustrate the robustness, efficiency and the rapidity of the proposed framework.},

keywords = {Dense image matching, Plane detection, Point Cloud, Randomized Hough transform},

pubstate = {published},

tppubtype = {article}

}