Final Report Summary - SURF3DSLAM (Probabilistic 3D surface matching for bathymetry based Simultaneous Localization and Mapping of underwater vehicles)
The project entitled “Probabilistic 3D surface matching for bathymetry based Simultaneous Localization and Mapping of underwater vehicles (Surf3DSLAM)”, financed through a Marie Curie European Reintegration Grant (Call: FP7-PEOPLE-2010-RG), has been executed from 1st April 2011 till 31th March 2013 by Mr. Angelos Mallios at the Universitat de Girona (Girona, Spain). The goal of this project was the development of an innovative technique for sonar-based Simultaneous Localization and Mapping (SLAM) with an Autonomous Underwater Vehicle (AUV), using bathymetry information gathered from a multi-beam sonar profiler. The primary objective of this research project was to extend to 3D the 2D technique for probabilistic, sonar-based, range-image registration that had initially developed from Mr. Mallios, and to use it for bathymetric pose-based SLAM.
The most relevant results were published this year (Probabilistic Surface Matching for Bathymetry Based SLAM, ICRA 2013). In that work we described a probabilistic surface matching method for pose-based bathymetry SLAM using a multibeam sonar profiler. The proposed algorithm compounds swath profiles of the seafloor with dead reckoning localization to build surface patches. Then, a probabilistic implementation of the ICP is used to deal with the uncertainty of the robot pose as well as the measured points in a two-stage process including point-to-point and point-to-plane metrics. A novel surface adaptation using octrees is proposed to have ICP-derived methods working in feature-poor or highly unstructured areas typical of bathymetric scenarios. Moreover, a heuristic based on the uncertainties of the surface points is used to improve the basic algorithm, decreasing the ICP complexity to O(n). The performance of the method was demonstrated with real data from a bathymetric survey.
In addition, we investigated a different method for a new surface adaptation based on the difference of normals as a multi-scale operator in unorganized point clouds and octrees (Bathymetry-based SLAM with Difference of Normals Point-Cloud subsampling and probabilistic ICP registration, OCEANS 2013).
The research above has benefitted from the experience of Mr. Mallios with his work on the 2D probabilistic scan matching SLAM, which was the base for the new 3D algorithms of this project. This knowledge has been transferred to the present project. Thus, the present project also includes the results of his work in navigation in a natural underwater environment (Navigating and Mapping with the Sparus AUV in a Natural and Unstructured Underwater Environment, OCEANS 2011, IEEE Best student poster award).
The research subjects treated in the present project have been proven to be significant for the robotics research and in oceanography research as well. It is important to mention the interdisciplinary scope of this research. Precise localization of an AUV is strong interdependent with many research lines in robotics and ocean science. Mission planning and control, real-time mission control, obstacle avoidance and multiple vehicle control are some of the robotics research lines where navigation is an essential input. In ocean sciences, AUVs are employed for small and large scale surveys. As a few examples, oceanographers collecting data of physical aspects from the water column for global climate change research. In marine geosciences, AUVs are building bathymetric maps on large scale surveys whilst in underwater archaeology are collecting thousand of pictures for a photo-mosaic reconstruction of the site. But AUVs are not used only for science. Harbor constructions, offshore industry and most important, search and rescue operations are benefiting more and more with the AUVs advances.
To achieve interdisciplinary needs, the project has benefitted from both the experience of Mr. Mallios as a Marie Curie ESR Fellow 2007-10, as well as the research carried out in the host organization.
Overall, the research carried out by Mr. Mallios with the host research group along these two years under the European Reintegration Grant has conducted to 12 peer-reviewed publications, one book chapter and two posters (one gained IEEE Best student poster award) that presented in different international conferences.
The most relevant results were published this year (Probabilistic Surface Matching for Bathymetry Based SLAM, ICRA 2013). In that work we described a probabilistic surface matching method for pose-based bathymetry SLAM using a multibeam sonar profiler. The proposed algorithm compounds swath profiles of the seafloor with dead reckoning localization to build surface patches. Then, a probabilistic implementation of the ICP is used to deal with the uncertainty of the robot pose as well as the measured points in a two-stage process including point-to-point and point-to-plane metrics. A novel surface adaptation using octrees is proposed to have ICP-derived methods working in feature-poor or highly unstructured areas typical of bathymetric scenarios. Moreover, a heuristic based on the uncertainties of the surface points is used to improve the basic algorithm, decreasing the ICP complexity to O(n). The performance of the method was demonstrated with real data from a bathymetric survey.
In addition, we investigated a different method for a new surface adaptation based on the difference of normals as a multi-scale operator in unorganized point clouds and octrees (Bathymetry-based SLAM with Difference of Normals Point-Cloud subsampling and probabilistic ICP registration, OCEANS 2013).
The research above has benefitted from the experience of Mr. Mallios with his work on the 2D probabilistic scan matching SLAM, which was the base for the new 3D algorithms of this project. This knowledge has been transferred to the present project. Thus, the present project also includes the results of his work in navigation in a natural underwater environment (Navigating and Mapping with the Sparus AUV in a Natural and Unstructured Underwater Environment, OCEANS 2011, IEEE Best student poster award).
The research subjects treated in the present project have been proven to be significant for the robotics research and in oceanography research as well. It is important to mention the interdisciplinary scope of this research. Precise localization of an AUV is strong interdependent with many research lines in robotics and ocean science. Mission planning and control, real-time mission control, obstacle avoidance and multiple vehicle control are some of the robotics research lines where navigation is an essential input. In ocean sciences, AUVs are employed for small and large scale surveys. As a few examples, oceanographers collecting data of physical aspects from the water column for global climate change research. In marine geosciences, AUVs are building bathymetric maps on large scale surveys whilst in underwater archaeology are collecting thousand of pictures for a photo-mosaic reconstruction of the site. But AUVs are not used only for science. Harbor constructions, offshore industry and most important, search and rescue operations are benefiting more and more with the AUVs advances.
To achieve interdisciplinary needs, the project has benefitted from both the experience of Mr. Mallios as a Marie Curie ESR Fellow 2007-10, as well as the research carried out in the host organization.
Overall, the research carried out by Mr. Mallios with the host research group along these two years under the European Reintegration Grant has conducted to 12 peer-reviewed publications, one book chapter and two posters (one gained IEEE Best student poster award) that presented in different international conferences.