Sistemi riconfigurabili a basso consumo per applicazioni di monitoraggio distribuito

The term Ambient Intelligence (AmI) refers to a vision on the future of the information society where smart, electronic environment are sensitive and responsive to the presence of people and their activities (Context awareness). In an ambient intelligence world, devices work in concert to support people in carrying out their everyday life activities, tasks and rituals in an easy, natural way using information and intelligence that is hidden in the network connecting these devices. This promotes the creation of pervasive environments improving the quality of life of the occupants and enhancing the human experience. AmI stems from the convergence of three key technologies: ubiquitous computing, ubiquitous communication and natural interfaces. Ambient intelligent systems are heterogeneous and require an excellent cooperation between several hardware/software technologies and disciplines, including signal processing, networking and protocols, embedded systems, information management, and distributed algorithms. Since a large amount of fixed and mobile sensors embedded is deployed into the environment, the Wireless Sensor Networks is one of the most relevant enabling technologies for AmI. WSN are complex systems made up of a number of sensor nodes which can be deployed in a target area to sense physical phenomena and communicate with other nodes and base stations. These simple devices typically embed a low power computational unit (microcontrollers, FPGAs etc.), a wireless communication unit, one or more sensors and a some form of energy supply (either batteries or energy scavenger modules). WNS promises of revolutionizing the interactions between the real physical worlds and human beings. Low-cost, low-computational power, low energy consumption and small size are characteristics that must be taken into consideration when designing and dealing with WSNs. To fully exploit the potential of distributed sensing approaches, a set of challengesmust be addressed. Sensor nodes are inherently resource-constrained systems with very low power consumption and small size requirements which enables than to reduce the interference on the physical phenomena sensed and to allow easy and low-cost deployment. They have limited processing speed,storage capacity and communication bandwidth that must be efficiently used to increase the degree of local ”understanding” of the observed phenomena. A particular case of sensor nodes are video sensors. This topic holds strong interest for a wide range of contexts such as military, security, robotics and most recently consumer applications. Vision sensors are extremely effective for medium to long-range sensing because vision provides rich information to human operators. However, image sensors generate a huge amount of data, whichmust be heavily processed before it is transmitted due to the scarce bandwidth capability of radio interfaces. In particular, in video-surveillance, it has been shown that source-side compression is mandatory due to limited bandwidth and delay constraints. Moreover, there is an ample opportunity for performing higher-level processing functions, such as object recognition that has the potential to drastically reduce the required bandwidth (e.g. by transmitting compressed images only when something ‘interesting‘ is detected). The energy cost of image processing must however be carefully minimized. Imaging could play and plays an important role in sensing devices for ambient intelligence. Computer vision can for instance be used for recognising persons and objects and recognising behaviour such as illness and rioting. Having a wireless camera as a camera mote opens the way for distributed scene analysis. More eyes see more than one and a camera system that can observe a scene from multiple directions would be able to overcome occlusion problems and could describe objects in their true 3D appearance. In real-time, these approaches are a recently opened field of research. In this thesis we pay attention to the realities of hardware/software technologies and the design needed to realize systems for distributed monitoring, attempting to propose solutions on open issues and filling the gap between AmI scenarios and hardware reality. The physical implementation of an individual wireless node is constrained by three important metrics which are outlined below. Despite that the design of the sensor network and its sensor nodes is strictly application dependent, a number of constraints should almost always be considered. Among them: • Small form factor to reduce nodes intrusiveness. • Low power consumption to reduce battery size and to extend nodes lifetime. • Low cost for a widespread diffusion. These limitations typically result in the adoption of low power, low cost devices such as low powermicrocontrollers with few kilobytes of RAMand tenth of kilobytes of program memory with whomonly simple data processing algorithms can be implemented. However the overall computational power of the WNS can be very large since the network presents a high degree of parallelism that can be exploited through the adoption of ad-hoc techniques. Furthermore through the fusion of information from the dense mesh of sensors even complex phenomena can be monitored. In this dissertation we present our results in building several AmI applications suitable for a WSN implementation. The work can be divided into two main areas:Low Power Video Sensor Node and Video Processing Alghoritm and Multimodal Surveillance . Low Power Video Sensor Nodes and Video Processing Alghoritms In comparison to scalar sensors, such as temperature, pressure, humidity, velocity, and acceleration sensors, vision sensors generate much higher bandwidth data due to the two-dimensional nature of their pixel array. We have tackled all the constraints listed above and have proposed solutions to overcome the current WSNlimits for Video sensor node. We have designed and developed wireless video sensor nodes focusing on the small size and the flexibility of reuse in different applications. The video nodes target a different design point: the portability (on-board power supply, wireless communication), a scanty power budget (500mW),while still providing a prominent level of intelligence, namely sophisticated classification algorithmand high level of reconfigurability. We developed two different video sensor node: The device architecture of the first one is based on a low-cost low-power FPGA+microcontroller system-on-chip. The second one is based on ARM9 processor. Both systems designed within the above mentioned power envelope could operate in a continuous fashion with Li-Polymer battery pack and solar panel. Novel low power low cost video sensor nodes which, in contrast to sensors that just watch the world, are capable of comprehending the perceived information in order to interpret it locally, are presented. Featuring such intelligence, these nodes would be able to cope with such tasks as recognition of unattended bags in airports, persons carrying potentially dangerous objects, etc.,which normally require a human operator. Vision algorithms for object detection, acquisition like human detection with Support Vector Machine (SVM) classification and abandoned/removed object detection are implemented, described and illustrated on real world data. Multimodal surveillance: In several setup the use of wired video cameras may not be possible. For this reason building an energy efficient wireless vision network for monitoring and surveillance is one of the major efforts in the sensor network community. Energy efficiency for wireless smart camera networks is one of the major efforts in distributed monitoring and surveillance community. For this reason, building an energy efficient wireless vision network for monitoring and surveillance is one of the major efforts in the sensor network community. The Pyroelectric Infra-Red (PIR) sensors have been used to extend the lifetime of a solar-powered video sensor node by providing an energy level dependent trigger to the video camera and the wireless module. Such approach has shown to be able to extend node lifetime and possibly result in continuous operation of the node.Being low-cost, passive (thus low-power) and presenting a limited form factor, PIR sensors are well suited for WSN applications. Moreover techniques to have aggressive power management policies are essential for achieving long-termoperating on standalone distributed cameras needed to improve the power consumption. We have used an adaptive controller like Model Predictive Control (MPC) to help the system to improve the performances outperforming naive power management policies.

A novel adaptive weight selection algorithm for multi-objective multi-agent reinforcement learning

To solve multi-objective problems, multiple reward signals are often scalarized into a single value and further processed using established single-objective problem solving techniques. While the field of multi-objective optimization has made many advances in applying scalarization techniques to obtain good solution trade-offs, the utility of applying these techniques in the multi-objective multi-agent learning domain has not yet been thoroughly investigated. Agents learn the value of their decisions by linearly scalarizing their reward signals at the local level, while acceptable system wide behaviour results. However, the non-linear relationship between weighting parameters of the scalarization function and the learned policy makes the discovery of system wide trade-offs time consuming. Our first contribution is a thorough analysis of well known scalarization schemes within the multi-objective multi-agent reinforcement learning setup. The analysed approaches intelligently explore the weight-space in order to find a wider range of system trade-offs. In our second contribution, we propose a novel adaptive weight algorithm which interacts with the underlying local multi-objective solvers and allows for a better coverage of the Pareto front. Our third contribution is the experimental validation of our approach by learning bi-objective policies in self-organising smart camera networks. We note that our algorithm (i) explores the objective space faster on many problem instances, (ii) obtained solutions that exhibit a larger hypervolume, while (iii) acquiring a greater spread in the objective space.

Analisi e applicazione dei processi di data mining al flusso informativo di sistemi real-time: Adattamento di un algoritmo di apprendimento automatico per la caratterizzazione e la ricerca di frequent patterns su macchine automatiche

La tesi da me svolta durante questi ultimi sei mesi è stata sviluppata presso i laboratori di ricerca di IMA S.p.a.. IMA (Industria Macchine Automatiche) è una azienda italiana che naque nel 1961 a Bologna ed oggi riveste il ruolo di leader mondiale nella produzione di macchine automatiche per il packaging di medicinali. Vorrei subito mettere in luce che in tale contesto applicativo l’utilizzo di algoritmi di data-mining risulta essere ostico a causa dei due ambienti in cui mi trovo. Il primo è quello delle macchine automatiche che operano con sistemi in tempo reale dato che non presentano a pieno le risorse di cui necessitano tali algoritmi. Il secondo è relativo alla produzione di farmaci in quanto vige una normativa internazionale molto restrittiva che impone il tracciamento di tutti gli eventi trascorsi durante l’impacchettamento ma che non permette la visione al mondo esterno di questi dati sensibili. Emerge immediatamente l’interesse nell’utilizzo di tali informazioni che potrebbero far affiorare degli eventi riconducibili a un problema della macchina o a un qualche tipo di errore al fine di migliorare l’efficacia e l’efficienza dei prodotti IMA. Lo sforzo maggiore per riuscire ad ideare una strategia applicativa è stata nella comprensione ed interpretazione dei messaggi relativi agli aspetti software. Essendo i dati molti, chiusi, e le macchine con scarse risorse per poter applicare a dovere gli algoritmi di data mining ho provveduto ad adottare diversi approcci in diversi contesti applicativi: • Sistema di identificazione automatica di errore al fine di aumentare di diminuire i tempi di correzione di essi. • Modifica di un algoritmo di letteratura per la caratterizzazione della macchina. La trattazione è così strutturata: • Capitolo 1: descrive la macchina automatica IMA Adapta della quale ci sono stati forniti i vari file di log. Essendo lei l’oggetto di analisi per questo lavoro verranno anche riportati quali sono i flussi di informazioni che essa genera. • Capitolo 2: verranno riportati degli screenshoot dei dati in mio possesso al fine di, tramite un’analisi esplorativa, interpretarli e produrre una formulazione di idee/proposte applicabili agli algoritmi di Machine Learning noti in letteratura. • Capitolo 3 (identificazione di errore): in questo capitolo vengono riportati i contesti applicativi da me progettati al fine di implementare una infrastruttura che possa soddisfare il requisito, titolo di questo capitolo. • Capitolo 4 (caratterizzazione della macchina): definirò l’algoritmo utilizzato, FP-Growth, e mostrerò le modifiche effettuate al fine di poterlo impiegare all’interno di macchine automatiche rispettando i limiti stringenti di: tempo di cpu, memoria, operazioni di I/O e soprattutto la non possibilità di aver a disposizione l’intero dataset ma solamente delle sottoporzioni. Inoltre verranno generati dei DataSet per il testing di dell’algoritmo FP-Growth modificato.

Un algoritmo di visione artificiale per l'individuazione dello switching point nel processo di etichettatura markerless

La tesi descrive il progetto e il funzionamento di un algoritmo basato su un sistema di visione, studiato allo scopo di eliminare i marker, che vengono utilizzati dai sensori di contrasto nel processo di taglio delle singole etichette a partire dalle bobine che scorrono sul rullo.

Analisi e applicazione dei processi di data mining al flusso informativo di sistemi real-time: Implementazione e analisi di un algoritmo autoadattivo per la ricerca di frequent patterns su macchine automatiche.

Analisi e applicazione dei processi di data mining al flusso informativo di sistemi real-time. Implementazione e analisi di un algoritmo autoadattivo per la ricerca di frequent patterns su macchine automatiche.

PhD forum : multiple camera management using wide base-line matching

Manual calibration of large and dynamic networks of cameras is labour intensive and time consuming. This is a strong motivator for the development of automatic calibration methods. Automatic calibration relies on the ability to find correspondences between multiple views of the same scene. If the cameras are sparsely placed, this can be a very difficult task. This PhD project focuses on the further development of uncalibrated wide baseline matching techniques.

A suspicious behaviour detection using a context space model for smart surveillance systems

Video surveillance systems using Closed Circuit Television (CCTV) cameras, is one of the fastest growing areas in the field of security technologies. However, the existing video surveillance systems are still not at a stage where they can be used for crime prevention. The systems rely heavily on human observers and are therefore limited by factors such as fatigue and monitoring capabilities over long periods of time. This work attempts to address these problems by proposing an automatic suspicious behaviour detection which utilises contextual information. The utilisation of contextual information is done via three main components: a context space model, a data stream clustering algorithm, and an inference algorithm. The utilisation of contextual information is still limited in the domain of suspicious behaviour detection. Furthermore, it is nearly impossible to correctly understand human behaviour without considering the context where it is observed. This work presents experiments using video feeds taken from CAVIAR dataset and a camera mounted on one of the buildings Z-Block) at the Queensland University of Technology, Australia. From these experiments, it is shown that by exploiting contextual information, the proposed system is able to make more accurate detections, especially of those behaviours which are only suspicious in some contexts while being normal in the others. Moreover, this information gives critical feedback to the system designers to refine the system.

Testing the accuracy of luminance maps acquired by smart phone cameras

This paper looks at the accuracy of using the built-in camera of smart phones and free software as an economical way to quantify and analyse light exposure by producing luminance maps from High Dynamic Range (HDR) images. HDR images were captured with an Apple iPhone 4S to capture a wide variation of luminance within an indoor and outdoor scene. The HDR images were then processed using Photosphere software (Ward, 2010.) to produce luminance maps, where individual pixel values were compared with calibrated luminance meter readings. This comparison has shown an average luminance error of ~8% between the HDR image pixel values and luminance meter readings, when the range of luminances in the image is limited to approximately 1,500cd/m2.

Survey report of NOAA Ship McArthur II cruises AR-04-04, AR-05-05 and AR-06-03: Habitat classification of side scan sonar imagery in support of deep-sea coral/sponge explorations at the Olympic Coast National Marine Sanctuary

Habitat mapping and characterization has been defined as a high-priority management issue for the Olympic Coast National Marine Sanctuary (OCNMS), especially for poorly known deep-sea habitats that may be sensitive to anthropogenic disturbance. As a result, a team of scientists from OCNMS, National Centers for Coastal Ocean Science (NCCOS), and other partnering institutions initiated a series of surveys to assess the distribution of deep-sea coral/sponge assemblages within the sanctuary and to look for evidence of potential anthropogenic impacts in these critical habitats. Initial results indicated that remotely delineating areas of hard bottom substrate through acoustic sensing could be a useful tool to increase the efficiency and success of subsequent ROV-based surveys of the associated deep-sea fauna. Accordingly, side scan sonar surveys were conducted in May 2004, June 2005, and April 2006 aboard the NOAA Ship McArthur II to: (1) obtain additional imagery of the seafloor for broader habitat-mapping coverage of sanctuary waters, and (2) help delineate suitable deep-sea coral/sponge habitat, in areas of both high and low commercial-fishing activities, to serve as sites for surveying-in more detail using an ROV on subsequent cruises. Several regions of the sea floor throughout the OCNMS were surveyed and mosaicked at 1-meter pixel resolution. Imagery from the side scan sonar mapping efforts was integrated with other complementary data from a towed camera sled, ROVs, sedimentary samples, and bathymetry records to describe geological and biological (where possible) aspects of habitat. Using a hierarchical deep-water marine benthic classification scheme (Greene et al. 1999), we created a preliminary map of various habitat polygon features for use in a geographical information system (GIS). This report provides a description of the mapping and groundtruthing efforts as well as results of the image classification procedure for each of the areas surveyed. (PDF contains 60 pages.)

Olympic Coast National Marine Sanctuary Habitat Mapping: Survey report and classification of side scan sonar data from surveys HMPR-114-2004-02 and HMPR-116-2005-01.

The Olympic Coast National Marine Sanctuary (OCNMS) continues to invest significant resources into seafloor mapping activities along Washington’s outer coast (Intelmann and Cochrane 2006; Intelmann et al. 2006; Intelmann 2006). Results from these annual mapping efforts offer a snapshot of current ground conditions, help to guide research and management activities, and provide a baseline for assessing the impacts of various threats to important habitat. During the months of August 2004 and May and July 2005, we used side scan sonar to image several regions of the sea floor in the northern OCNMS, and the data were mosaicked at 1-meter pixel resolution. Video from a towed camera sled, bathymetry data, sedimentary samples and side scan sonar mapping were integrated to describe geological and biological aspects of habitat. Polygon features were created and attributed with a hierarchical deep-water marine benthic classification scheme (Greene et al. 1999). For three small areas that were mapped with both side scan sonar and multibeam echosounder, we made a comparison of output from the classified images indicating little difference in results between the two methods. With these considerations, backscatter derived from multibeam bathymetry is currently a costefficient and safe method for seabed imaging in the shallow (<30 meters) rocky waters of OCNMS. The image quality is sufficient for classification purposes, the associated depths provide further descriptive value and risks to gear are minimized. In shallow waters (<30 meters) which do not have a high incidence of dangerous rock pinnacles, a towed multi-beam side scan sonar could provide a better option for obtaining seafloor imagery due to the high rate of acquisition speed and high image quality, however the high probability of losing or damaging such a costly system when deployed as a towed configuration in the extremely rugose nearshore zones within OCNMS is a financially risky proposition. The development of newer technologies such as intereferometric multibeam systems and bathymetric side scan systems could also provide great potential for mapping these nearshore rocky areas as they allow for high speed data acquisition, produce precisely geo-referenced side scan imagery to bathymetry, and do not experience the angular depth dependency associated with multibeam echosounders allowing larger range scales to be used in shallower water. As such, further investigation of these systems is needed to assess their efficiency and utility in these environments compared to traditional side scan sonar and multibeam bathymetry. (PDF contains 43 pages.)

A smart wireless inertial measurement unit system: simplifying & encouraging usage of WIMU technology

Wireless Inertial Measurement Units (WIMUs) combine motion sensing, processing & communications functionsin a single device. Data gathered using these sensors has the potential to be converted into high quality motion data. By outfitting a subject with multiple WIMUs full motion data can begathered. With a potential cost of ownership several orders of magnitude less than traditional camera based motion capture, WIMU systems have potential to be crucially important in supplementing or replacing traditional motion capture and opening up entirely new application areas and potential markets particularly in the rehabilitative, sports & at-home healthcarespaces. Currently WIMUs are underutilized in these areas. A major barrier to adoption is perceived complexity. Sample rates, sensor types & dynamic sensor ranges may need to be adjusted on multiple axes for each device depending on the scenario. As such we present an advanced WIMU in conjunction with a Smart WIMU system to simplify this aspect with 3 usage modes: Manual, Intelligent and Autonomous. Attendees will be able to compare the 3 different modes and see the effects of good andbad set-ups on the quality of data gathered in real time.

The smart vision local navigation aid for blind and visually impaired persons

The SmartVision prototype is a small, cheap and easily wearable navigation aid for blind and visually impaired persons. Its functionality addresses global navigation for guiding the user to some destiny, and local navigation for negotiating paths, sidewalks and corridors, with avoidance of static as well as moving obstacles. Local navigation applies to both in- and outdoor situations. In this article we focus on local navigation: the detection of path borders and obstacles in front of the user and just beyond the reach of the white cane, such that the user can be assisted in centering on the path and alerted to looming hazards. Using a stereo camera worn at chest height, a portable computer in a shoulder-strapped pouch or pocket and only one earphone or small speaker, the system is inconspicuous, it is no hindrence while walking with the cane, and it does not block normal surround sounds. The vision algorithms are optimised such that the system can work at a few frames per second.

Estudo e proposta de algoritmo para a minimização do corte de eólica no contexto de uma análise custo-benefício multicritério

Trabalho Final de Mestrado para obtenção do grau de Mestre em Engenharia Electrotécnica Ramo de Energia

Smart visible sets para ambientes de rede

A visualização em tempo real de cenas complexas através de ambientes de rede é um dos desafios na computação gráfica. O uso da visibilidade pré-computada associada a regiões do espaço, tal como a abordagem dos Potentially Visible Sets (PVS), pode reduzir a quantidade de dados enviados através da rede. Entretanto, o PVS para algumas regiões pode ainda ser bastante complexo, e portanto uma estratégia diferente para diminuir a quantidade de informações é necessária. Neste trabalho é introduzido o conceito de Smart Visible Set (SVS), que corresponde a uma partição das informações contidas no PVS segundo o ângulo de visão do observador e as distâncias entre as regiões. Dessa forma, o conceito de “visível” ou de “não-visível” encontrado nos PVS é estendido. A informação referente ao conjunto “visível” é ampliada para “dentro do campo de visão” ou “fora do campo de visão” e “longe” ou “perto”. Desta forma a informação referente ao conjunto “visível” é subdividida, permitindo um maior controle sobre cortes ou ajustes nos dados que devem ser feitos para adequar a quantidade de dados a ser transmitida aos limites impostos pela rede. O armazenamento dos SVS como matrizes de bits permite ainda uma interação entre diferentes SVS. Outros SVS podem ser adicionados ou subtraídos entre si com um custo computacional muito pequeno permitindo uma rápida alteração no resultado final. Transmitir apenas a informação dentro de campo de visão do usuário ou não transmitir a informação muito distante são exemplos dos tipos de ajustes que podem ser realizados para se diminuir a quantidade de informações enviadas. Como o cálculo do SVS depende da existência de informação de visibilidade entre regiões foi implementado o algoritmo conhecido como “Dual Ray Space”, que por sua vez depende do particionamento da cena em regiões. Para o particionamento da cena em uma BSP-Tree, foi modificada a aplicação QBSP3. Depois de calculada, a visibilidade é particionada em diferentes conjuntos através da aplicação SVS. Finalmente, diferentes tipos de SVS puderam ser testados em uma aplicação de navegação por um cenário 3D chamada BSPViewer. Essa aplicação também permite comparações entre diferentes tipos de SVS e PVS. Os resultados obtidos apontam o SVS como uma forma de redução da quantidade de polígonos que devem ser renderizados em uma cena, diminuindo a quantidade de informação que deve ser enviada aos usuários. O SVS particionado pela distância entre as regiões permite um corte rápido na informação muito distante do usuário. Outra vantagem do uso dos SVS é que pode ser realizado um ordenamento das informações segundo sua importância para o usuário, desde que uma métrica de importância visual tenha sido definida previamente.

A low cost one-camera optical tracking system for indoor wide-area augmented and virtual reality environments

In the last years the number of industrial applications for Augmented Reality (AR) and Virtual Reality (VR) environments has significantly increased. Optical tracking systems are an important component of AR/VR environments. In this work, a low cost optical tracking system with adequate attributes for professional use is proposed. The system works in infrared spectral region to reduce optical noise. A highspeed camera, equipped with daylight blocking filter and infrared flash strobes, transfers uncompressed grayscale images to a regular PC, where image pre-processing software and the PTrack tracking algorithm recognize a set of retro-reflective markers and extract its 3D position and orientation. Included in this work is a comprehensive research on image pre-processing and tracking algorithms. A testbed was built to perform accuracy and precision tests. Results show that the system reaches accuracy and precision levels slightly worse than but still comparable to professional systems. Due to its modularity, the system can be expanded by using several one-camera tracking modules linked by a sensor fusion algorithm, in order to obtain a larger working range. A setup with two modules was built and tested, resulting in performance similar to the stand-alone configuration.