Abstract
As the variety of mobile devices connected to the Internet grows, there is a corresponding increase in the need to deliver content tailored to their heterogeneous characteristics. At the same time, we watch to the increase of eLearning in universities through the adoption of electronic platforms and standards. Not surprisingly, the concept of mLearning (Mobile Learning) appeared in recent years decreasing the limitation of learning location with the mobility of general portable devices. However, this large number and variety of Web-enabled devices poses several challenges for Web content creators who want to automatic get the delivery context and adapt the content to the client mobile devices. In this paper we analyze several approaches to defining delivery context and present an architecture for deliver uniform mLearning content to mobile devices denominated eduMCA-Educational Mobile Content Adaptation. With the eduMCA system the Web authors will not need to create specialized pages for each kind of device, since the content is automatically transformed to adapt to any mobile device capabilities from WAP to XHTML MP-compliant devices.

Abstract
This paper describes the interaction between the kinematic model of the AUV MARES and the measurement or observation of the environment through images obtained with a sonar. Three types of sonar are discussed in this paper: forward-look, side scan and multibeam - but the sonar used to develop this work was the side scan sonar. The type of observations and characteristics of the environment provided by the sonar are described here. The method which connects the sensory part of the vehicle with the observations from the sonar, was the Kalman filter (EKF). In this paper, we present two simulations of filters for two different characteristics. Both filters estimate the characteristics of the natural landmarks, creating an environment map, but both of them consider different states of the vehicle. Results of the simulation are obtained. The features that are considered are an underwater pipe on the floor and a vertical wall. A control loop for the vehicle that provides the capacity to move along the feature/landmark from a reference distance is also discussed.

Abstract
The segmentation of satellite images is a valuable tool to obtain useful information about the spatial distribution of different land cover types. The use of segmentation algorithms instead of the traditional pixel-by-pixel classifiers used to produce land cover maps results on images that exhibit a more homogeneous distribution of classes, showing the piecewise spatial continuity of the real world. Several segmentation and classification methods are being developed to properly handle the high dimensionality of hyperspectral images. An example is a Bayesian segmentation procedure based on discriminative classifiers with a Multi-Level Logistic Markov-Gibbs prior. This method adopts the Fast Sparse Multinomial Logistic Regression as discriminative classifier, a method that promotes sparsity by including a Laplacian prior. However, the use of this type of prior requires an extensive search to for the best parameter of sparsity. In this work, a modification to this method is introduced. Instead of using the Laplacian Prior to enforce the sparsity of FSMLR classifier, the Jeffreys prior is used. This prior avoids the need to proceed to an extensive search for the best parameter, and also keeps the sparsity of the densities estimators, resulting on a faster and competitive segmentation procedure. The results of the application of this new approach to the benchmarked dataset Indian Pines show the effectiveness of the proposed method when compared with that using the Laplacian prior.

Abstract
According to the World Health Organization in 2002 more than 161 million people were visually impaired [1]. The mobility of the visually impaired reveals itself to be a rather complex task, even with the support of inclusive technologies. Currently, the most widespread and used means by the visually impaired people are the white stick and the guide dog, however both show some difficulties in their mobility. The white stick only detects obstacles at the ground level, whilst obtaining a guide dog correctly trained in Portugal is an expensive and slow solution [2], given that the communication between the visually impaired person and the dog is limited. If the blind person had access to reliable information, and in real time, about the environment in which he lives, it certainly would be safer for him to move around it [3]. Thus, more and more, with the increasing development of the technologies, there is the possibility of creating systems that assist the visually impaired people in their task of detecting the obstacles and subsequent orientation. The objective of this article is to create a system that allows the capture of images in real time and that accomplishes the segmentation resorting to the information with disparity. Later on, the result of the image segmentation will be used in a system that offers the visually impaired person information on the environment that surrounds him. The image acquisition is done through a system of stereo vision that captures images from the surrounding environment as the user moves around.

Abstract
Organizations with a presence in the Second Life (R) world typically only provide direct user interaction with staff at specific schedules, or not at all. We present a system that provides organizations with a simple way to enable constant interaction with users of the Second Life world, by simulating staff presence using automated avatars as communication channels to real-life staff by means of instant messaging and short message service technologies. Staff members are assigned to communication with Second Life avatars based on a hierarchy of information desk staffing priorities, and communication is bidirectional.

Abstract
In the robotic domain, it is common to deduce and use models that allow translating mathematically the element behavior. In some cases, these would serve as base to determine and develop a controller, for example. Beyond this, the simulation and experiments are reasons that leave to the development of models, becoming evaluation tools of the system behavior, especially when there are constraints of implementation or in experiments. However, the modeling is an approach to the reality, since it is difficult to translate the behavior of an element in a strict way and the disturbances to witch it is subject to. In this work, we address the modeling questions of an autonomous underwater vehicle. This paper describes the deducing of a dynamic model with six degrees of freedom of an underwater vehicle, considering all of its physical characteristics. This is achieved by the determination of all forces that actuates on the body during its motions and by the determination of the rigid body dynamic. The modeling method is presented as well as the coefficients determination. Finally, a comparison with experimental results is carried out.