Abstract
The growing economic interest in the exploitation of mineral resources on deep-ocean beds, including those in the vicinity of sensitive-rich habitats such as hydrothermal vents, raise a mounting concern about the damage that such actions might originate to these poorly-know ecosystems, which represent millions of years of evolution and adaptations to extreme environmental conditions. It has been suggested that mining may cause a major impact on vent ecosystems and other deep-sea areas. Yet, the scale and the nature of such impacts are unknown at present. Hence, building upon currently available scientific information it is crucial to develop new cost-effective technologies embedded into rigorous operating frameworks. The forward-thinking provided here will assist in the development of new technologies and tools to address the major challenges associated with deep sea-mining; technologies for in situ and ex situ observation and data acquisition, biogeochemical processes, hazard assessment of deep-sea mining to marine organisms and development of modeling tools in support of risk assessment scenarios. These technological developments are vital to validate a responsible and sustainable exploitation of the deep-sea mineral resources, based on the precautionary principle. © 2018 Elsevier B.V.

Abstract
The use of the odometry and SLAM visual methods in autonomous vehicles has been growing. Optical sensors provide valuable information from the scenario that enhance the navigation of autonomous vehicles. Although several visual techniques are already available in the literature, their performance could be significantly affected by the scene captured by the optical sensor. In this context, this paper presents a comparative analysis of three monocular visual odometry methods and three stereo SLAM techniques. The advantages, particularities and performance of each technique are discussed, to provide information that is relevant for the development of new research and novel robotic applications. © Springer International Publishing AG 2018.

Abstract

Abstract
This paper addresses the use of a hyperspectral image system to detect vessels in maritime operational scenarios. The developed hyperspectral imaging classification methods are based on supervised approaches and allow to detect the presence of vessels using real hyperspectral data. We implemented two different methods for comparison purposes: SVM and SAM. The SVM method, which can be considered one of most utilized methods for image classification, was implemented using linear, RBF, sigmoid and polynomial kernels with PCA for dimensionality reduction, and compared with SAM using a two classes definition, namely vessel and water. The obtained results using real data collected from a UAV allow to conclude that the SVM approach is suitable for detecting the vessel presence in the water with a precision and recall rates favorable when compared to SAM. © 2018 IEEE.

Abstract