Innovative Eti Technologies

Transcript

The Innovative Faculty for Innovative Technologies Gdańsk University of Technology Faculty of Electronics, Telecommunications and Informatics www.eti.pg.gda.pl T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 1 The Faculty at a glance • • • • • • • • 16 departments 4 fields of studies 3600 students 500 graduates each year 20 doctorates each year 200 researchers, incl. 20 Full Professors 50 running research projects 125 cooperating companies Laboratories • • • • • • • • • • • • • • • • • • • • • • • • • • Algorithms and System Modelling Lab Computer Architecture Lab Biomedical Engineering Lab Microwave and Antenna Engineering Lab Software and Knowledge Engineering Lab Intelligent Interactive Systems Lab Decision Systems Lab Metrology Lab and Optoelectronics Lab Automatic Control Lab Marine Electronic Systems Lab Geoinformatics Lab Computer Communications Lab Microelectronic Systems Lab Multimedia Systems Lab Radiocommunication Systems and Networks Lab Teleinformation Networks Lab Anechoic Chambers Advanced Computer Animations Lab Industrial Robots Laboratory Innovative Materials and Devices Lab (Clean-room) Ground Satellite Station Lab Computer Networks Lab Mobile Devices Lab Immersive 3D Graphics Lab (under construction) Displays and Colorimetry Lab Optical Coherence Tomography Lab Participation in European Research Area • • • • • • 2 FP5,FP6, FP7 Tempus COST Norway Funds Eureka Artemis T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s The Faculty of Electronics, Telecommunications and Informatics (ETI) educates about 3,500 students in four fields of studies: Informatics, Electronics and Telecommunications (considered to be one field), Control Engineering, Robotics, and Biomedical Engineering. We are the largest faculty in Northern Poland in the area of modern Telecommunications and Information Technologies employing nearly 200 researchers and lecturers, which includes over 40 Professors and Doctors of Science. For many years we have retained a high scientific standing, documented by the highest category “A” according to the Ministry of Science and Higher Education’s classification system, and by being entitled to confer a DSc degree in three disciplines: Electronics, Telecommunications and Informatics as well as a PhD degree in the discipline of Automatics and Robotics. The Faculty has an extensive research and didactic infrastructure at its disposal. Recently, a new (and the Faculty’s second) building was made available, serving primarily didactic purposes – it houses laboratories, lecture and auditorium halls, comprising a total of about 1550 seats, as well as a student-friendly Faculty office. There is also the Tri-City Academic Computer Centre, called by its Polish acronym “TASK”, with Galera+, one of the fastest supercomputers in Europe, as well as a State-of-the-Art library and reading room. There are many extensively equipped research and didactic laboratories. Two anechoic chambers are used for acoustic research and antenna measurements. An (one of the few in Poland) earth satellite station for research is installed on the roof of Building A of the ETI faculty. A unique laboratory of industrial robots was opened last year, and soon the construction of premises for an Immersive 3D Graphics Laboratory will be commenced. This laboratory will be unique on the European scale and is designed to develop interdisciplinary research on virtual reality. T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 3 Graduates of the Faculty have established successful world-wide known companies. Just two prominent examples are: Young Digital Planet, which is a world leader in the field of digital educational content, and IVO Software (now owned by Amazon.com), who are inventors of the international award-winning speech synthesizer “Ivona”. A large number of the Faculty’s graduates have been very successful abroad. The most prominent example is Prof. Tomasz Imieliński, the world’s most cited Polish computer scientist, currently residing in the USA. The Faculty of Electronics, Telecommunications and Informatics at Gdańsk University of Technology is a modern academic centre harmoniously combining advanced research and attractive education with active cooperation in the industrial environment. Within the framework of the latter, the Faculty initiated the formation of the Pomeranian ICT Cluster, which integrates over a hundred business partners from the fields of Information, Electronics and Telecommunications technologies in their broader sense, as well as several educational institutions and local government organizations. The inventions developed in our Faculty are widely known and recognized. Among them, there are computer tools for hearing, speech and sight screening tests, as well as a range of dedicated specialized interfaces enabling paralyzed and handicapped people to use computers. An important group are the inventions that support 4 therapy in the case of common conditions, such as speech impairment or circulatory disorders. Other inventions of great practical importance are: a passive acoustic radar, a series of mobile impedance analyzers for the diagnostics of anti-corrosion coatings, a unique mobile expert system for measuring the motion of gaseous pollutants in the atmospheric air, a noise monitoring system, a system for detecting trace amounts of hazardous substances, or, a new (recently awarded with the trademark “Poland Now!” label) system of smart LED lamps. In this booklet you will find short descriptions of some of the inventions mentioned above, as well as the newest one, developed just recently. The Faculty of Electronics, Telecommunications and Informatics, is pleased to be a modern and fast-developing scientific and didactic unit of such a modern technical institution as Gdańsk University of Technology. We invite you to any kind of benign cooperation to meet the great challenge - the harmonious and dynamic development of the 21st century Society. Prof. Krzysztof Goczyła Dean of the Faculty T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Active Dynamic Thermography for quantitative skin burn diagnostics Purpose The concept of using thermal excitation and measurement of thermal responses is known as Non Destructive Thermal Testing (NDTT). Active Dynamic Thermography (ADT) allows non-destructive testing of bulky, multi-structure, foil or composite materials using external heat excitation and visualisation of the infrared response, showing the surface temperature distribution and thermal transients. Our investigations are directed mainly towards medical applications and development of new methods of diagnostics, e.g.: non-contact evaluation of the degree of skin burn damage; evaluation of treatment procedures; evaluation of intra-operation procedures in cardiosurgery etc. The aim of the work was to find the relationship between figures of merit in static thermography ΔT (difference of mean values of the burn wound and of not-affected skin area temperatures) and in ADT (thermal time constants parametric images - tau) for burn classification most suitable for choosing treatment. Statistical analysis did show a high correlation between parameters: thermal time constants, ΔT, the histopathology estimation and the burn wound classification into healed and unhealed during three weeks. The study results revealed the quantitative criterions tau and ΔT for burn classification allowing a proper choice of burn wound treatment. Description Capturing of IR-images while cooling or heating allows for calculation of thermal images used for classification of regions for conservative or surgical treatment. This figure shows ADT set with cooling applied for burn diagnostics in clinics. Contact Prof. Antoni Nowakowski Department of Biomedical Engineering phone: +48 58 347 26 45, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 5 AEGIR – Asynchronous, Autonomous and Ground-Based Radiolocation System Purpose The AEGIR system is designed to be an independent radiolocation system, that will allow you to determine locators’ positions without using any satellite navigation systems (like GPS or GLONASS systems). Description In the AEGIR system all ground reference stations operate in an asynchronous way, so each one uses a local oscillator to transmit localization messages. They can also receive signals from neighboring stations. Based on the received signals, a reference station determines the time difference between the neighboring reference stations and its own rhythm of work. The measurement results are regularly placed in the localization message. The locator receiver is responsible for the synchronization process. On the basis of its own measurements and measurements from the reference station, it estimates its location. Compared to existing solutions like Loran-C (Long Range Navigation - C), the AEGIR system resigns from a chain relationship between reference stations. In the proposed system, there are no supervision centers for maintenance, which reduces operating costs and increases system reliability. The AEGIR system will be fully functional if the locator will receive signals from at least three reference stations (2D case) or four reference stations (3D case). The principle of the system is under patent protection in this country and in the European Union. In addition, the system has the advantage that in the process of estimating the position, coordinates are determined directly, without the need to assign them to the so-called positioning lines that are needed in phase radiolocation systems. Therefore, there is no need to input a starting position and count the excess position lines that are repeated on a regular basis in the area of the system, and that can lead to errors in determining the correct latitude and longitude. AEGIR - in Norse mythology: god of ocean Contact Prof. Jacek Stefański, PhD, DSc Department of Radio Communication Systems and Networks phone: +48 58 347 15 66, email: [email protected] 6 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s An Analog Sub-Milliwatt CMOS Image Sensor With Pixel-Level Convolution Processing Purpose The developed image sensor with vision processors is integrated on a single chip to create a so-called vision chip. Such kinds of chips find many applications in robotics, biomedical implants, road traffic control systems, and navigation systems. The main function of the vision chips is to perform numerically-intensive early vision processing algorithms, including smoothing, edge detection, noise reduction, sharpening, etc. The low-power operation of such chips is of great importance for many autonomous systems powered from batteries or small solar cells. Description A new approach to an analog ultra-low power medium-resolution vision chip design has been invented. The prototype chip performs low-level image processing algorithms in real-time. Only a photo-diode, MOS switches and two capacitors are used to create an analog processing element (APE) that is able to realize any convolution algorithm based on a full 3 x 3 kernel. The proof-of-concept circuit was fabricated in 0.35 µm CMOS technology, and contains a 64 x 64 single instruction multiple data (SIMD) matrix with embedded APEs. The prototype has been functionally tested and measured. The performed tests show that the vision chip can be used for convolution image filtering The matrix dissipates less than 0.3 mW (less than 0.1 µW per APE) of power under 3.3 V supply, and its image processing speed is up to 100 frames/s. Main chip parameters Technology Supply Array size Chip area APE size Power per APE Frame rate Random noise Optical dynamic range Linearity 0.35 µm CMOS 3.3V 64 x 64 9.8 mm2 35 µm x 35 µm 0.07 µW max 10-100 fps 5mVrms @ 500lx 58 dB 97 % Contact Prof. Stanisław Szczepański Department of Microelectronic Systems phone: +48 58 347 22 78, fax: +48 58 347 23 78, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 7 Auditory-visual attention stimulator Purpose It is estimated that about 2 % of children have undermined lateralization of the brain. The proposed solution is based on the assumption that lateralization of sight/hearing could be focused by the stimulation of those senses in an appropriate way. Features of the system will allow for the correction of, and consistent development of, children; especially in the case of children with various impairments (especially cross-dominance or dyslexia). Description The main idea is to perform parallel stimulation of the sight and hearing senses using digital signal processing techniques. Modification of visual and hearing stimuli is performed in order to focus perception of those senses by the appropriate hemisphere (by means of the lateralization profile). Sense of hearing is stimulated using the time scale modification (TSM) of speech combined with amplitude modification of speech used during the training. The sight is stimulated by displaying the text of the spoken elements. Text is modified during the training, i.e. part of the text that is currently heard in the headphones is marked using various techniques (e.g. zooming, colouring, etc.). In the extended version of the system, parameters associated with vision and hearing modification algorithms might be controlled by a gaze tracker. Contact Dr. Piotr Odya, PhD Multimedia Systems Department phone: +48 58 347 23 01, fax: +48 58 347 11 14, email: [email protected] 8 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Automatic video event detection software Purpose This product is the prototype of a video processing and event detection solution designed to be applied in multi-camera surveillance systems. It detects moving objects in fixed camera streams, tracks them, classifies them, and detects user-defined events including intrusion and wrong-direction movement, as well as specialized events, such as abandoned luggage, events occurring in a parking lot (parking time and place identification, entering and leaving a parking lot, etc.), and traffic events (prohibited lane change, stopping in the middle of an intersection, etc.). All events are instantaneously reported to the system operator. Description The system can operate on any Windows or Linux OS PCs with digital IP cameras attached. The system is highly distributed and scalable, therefore multiple servers may be employed to the processing of streams from a large number of cameras. Results of video analysis are transmitted to dedicated end user terminals by wired links, WiFi, or GPRS, including PCs and Smartphones. Intrusion detection example Prohibited lane change detection Abandoned luggage detection Contact Dr. Piotr Szczuko, PhD Multimedia Systems Department phone: +48 58 347 29 72, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 9 Hybrid Localization System Purpose The system’s aim is to localize active RFID tags indoors, in a more effective way than traditional systems. Localization methods can use different environmental features (signals) and their properties: different characteristics associated with the propagation of radio signals (RSS, ToF, DoA, etc.), data recorded from sensors on a tag (infrared, compass, accelerometers, ultrasound, etc.), images recorded by a video cameras (computer vision). The Hybrid Localization System relies on measurements of various signals from the environment and different signal properties, uses advanced algorithms for data processing, so different components of the system can mutually remove their weak points and improve the overall accuracy whilst simultaneously keeping infrastructure costs low. Description Algorithms for hybrid localization Inexpensive switched beam antennas for AoA subsystems Inexpensive programmable cameras and Computer Vision algorithms supporting localization process in the hybrid system Contact Dr. Łukasz Kulas, PhD Department of Microwave and Antenna Engineering phone: +48 58 347 16 59 , email: [email protected] 10 This work has been supported by the Polish National Centre for Research and Development under agreement LIDER/23/147/L-1/09/NCBiR/2010. T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Indoor Radiolocation System for Special Applications SALOn Purpose The system was designed as a support for firemen or policemen during intervention in unknown buildings, to increase safety and reduce time-to-help in case of emergency Description SALOn is made up of three types of devices: • personal identification modules, worn by officers, • reference modules, installed around a/the supervised area/building, for example mounted on cars which are used by the officers to get to the area/building, • server responsible for data collection, calculation of officer’s positions and presentation (made of industrial PC). Positions of nodes (officers) in network are calculated on the basis of node-to-node distance measurements using radio waves and round-trip-time method. To avoid altitude ambiguity in case of all reference nodes mounted at equal height, an additional air pressure sensor is used to measure change in altitude by differences in measured pressure. Other kinds of sensors (magnetometers, inertial sensors) can be applied to increase the accuracy of location service. In addition to location data, some user data can be transmitted via the SALOn radio network: alarms triggered manually, alarms triggered by accelerometer (e.g. free fall detection), evacuation order etc. In case of an alarm message originated from one personal module, other users are informed via their modules about the I.D. and distance to that alarm initiator. The SALOn system was built as a working prototype with full functionality. A final version can be modified to fit a user’s requirements. Contact Dr. Jarosław Sadowski, PhD Department of Radio Communication Systems and Networks phone: +48 58 347 22 11, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 11 MedEye Purpose Supporting medical research, especially in gastrointestinal endoscopy Description The application deals with the analysis of the media stream that is the path of film from the endoscopic camera capsule. Endoscopy creates opportunities for rapid detection of neoplastic changes in gastrointestinal tracts. However, such research takes a long time and requires of the doctor a lot of concentration so that s/he can correctly identify the lesion. This application may be assisted by the investigator to keep pointing out the most important pieces of research, and to suggest the pathological changes that occur there. When playing a recording of the movie that is limited only to the most important pieces showing lesions, this significantly speeds up the decision-making. These operations can be performed on an ongoing basis, as well as remotely, using the KASKADA platform. The platform is deployed on a cluster supercomputer environment (the Galera supercomputer) which consists of multiple computational nodes connected by a fast low-latency network. It is capable of processing incoming data with the use of services created by the user, and provides methods for an easy parallelization of the algorithms. Contact Dr. Tomasz Dziubich, PhD Department of Computer Architecture phone: +48 58 347 25 24 email: [email protected] 12 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Multi component gas analyzer Purpose Low-cost, real time measurement solution for up to 10 environmental gasses’ detection and analysis. Description The analyzer is based on low-cost different type gas sensors. The device is an ideal solution for rapid screening of environmental pollutants. Unique measurement techniques and data analysis reduce cross-sensitivity and long term sensor drift, allowing long term operation without maintenance. The device has build logging capability and the possibility of remote control with GSM. Compact size and fast installation allow mobile or stationary application. The dedicated software allows communication with a few analyzers. Contact Dr. Grzegorz Jasiński, PhD Department of Biomedical Engineering phone: +48 58 347 17 35 email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 13 NOR-STA – Support for Achieving and Assessing Conformance to NORms and STAndards Purpose To deliver effective and efficient support for the processes of achieving, assessing, and maintaining conformance to norms and standards. Conformance is required in many sectors, and is among the priorities of Enterprise. Demonstrated conformance is commonly recognized as confirming the quality of delivered services and products. Description NOR-STA is an internet platform of innovative generic software services supporting application of norms and standards. Key benefits include: • arrangement of a standard’s requirements and the supporting evidence,, • maintenance of conformance, • support for conformance assessment and related decision processes, • cost and effort savings, and • security. NOR-STA supports all parties involved in the implementation of standards including owners, promoters, users, auditors and consultants. It provides a unique on-line and mobile communication platform with dedicated services. NOR-STA services are deployed in accordance with the Software-as-a-Service cloud computing model. The services have already been used to support some 50 customers in the implementation of a number of standards in different domains, including healthcare, outsourcing, public administration and education. Contact NOR-STA Portal: www.nor-sta.eu/en Project Manager: Prof. Janusz Górski phone: +48 58 347 19 09, email: [email protected] 14 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Object re-identification application for multi-camera surveillance systems Purpose The motivation to create this prototype is the need to identify the same object in video streams from different cameras characterized by various recording conditions. This method can be used for tracking objects moving from one camera to another, especially in case of overlapped Fields of View (FOVs). Defining of a common region in FOVs of a particular pair of cameras is a starting point to the solution of the problem of objects’ reidentification. For disjointed camera views a more sophisticated approach is necessary, because the video surveillance system can contain various types of cameras with different orientations, and illumination conditions can vary as well. The set of visual features and matching algorithms is applied in the elaborated software. Description The prototype works as part of a surveillance system equipped with non-overlapping cameras, making it possible to track moving objects in the case of discontinuous visual presence in the available set of cameras and under variable illumination conditions. Multi-camera object matching Variable visual features of objects Contact Prof. Andrzej Czyżewski, Mr. Karol Lisowski MSc. Multimedia Systems Department phone: +48 58 347 16 36, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 15 Object Tracking and Automatic Master-Slave PTZ Camera Positioning System Purpose This product is a prototype of object tracking and automatic pan-tilt-zoom (PTZ) camera positioning intended for multi-camera surveillance systems. It detects an object of interest in a fixed camera view, tracks its movement, and automatically controls a high zoom PTZ camera to follow the object. Objects leaving the camera’s area-of-coverage are automatically detected and recognized in nearby cameras, thus the tracking may be continued in larger areas. Description The application can operate on any Windows and Linux PC with digital IP cameras attached: at least one fixed, and one PTZ, camera are required. It is able to record video streams, to process them, and to transmit data to dedicated terminals by wired links, WiFi, and GPRS. Besides PC-based terminals, it supports also smartphones (with Windows Mobile OS), providing video streaming and touch screen operation for selecting the object of interest in the video. PC terminal with four FullHD displays Automatic PTZ camera tracking the object Contact Dr. Piotr Szczuko, PhD Multimedia Systems Department phone: +48 58 347 29 72, email: [email protected] 16 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Passive Acoustic Radar Purpose The Passive Acoustic Radar is a prototype solution that can be used for automatic detection, classification (4 classes: explosion, gunshot, scream, broken glass), localization, and tracking of sound sources, also employing a pan-tilt-zoom (PTZ) camera. Contrary to active radars, it does not emit any scanning beam but, after receiving environmental sounds, it provides information about the direction of an incoming acoustical signal. Therefore, if a hazardous sound event is detected (i.e. explosion, gunshot, broken glass, scream) the PTZ camera points automatically into the direction of the event. Description The device consists of some new kinds of multichannel miniature sound intensity sensors, and a set of digital signal processing algorithms. The functionality of the Passive Acoustic Radar enables: sound source visualization in a 3D space, and a precise PTZ camera steering towards the sound source direction. Multiple sound sources can also be detected and localized. Moreover, the implemented beamforming allows ones to obtain an „acoustical focus” on a part of the spatial scene, when the selected sound source needs to be monitored accurately. The proposed device can significantly improve the functionality of traditional surveillance monitoring systems. Passive Acoustic Radar in action Contact Dr. Józef Kotus, PhD Multimedia Systems Department phone: +48 58 347 29 72, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 17 Personalized Healthcare Solutions for Ambient Assisted Living Purpose Ubiquitous healthcare systems used as assisted living solutions will not only help to prevent, detect, and monitor health conditions but will also support sick, elderly, and disabled people to live independently. Health personalization and support for older and immobilized people is a very important target for many national and international initiatives. Description It is a multimodal, integrated platform for communication, training, and monitoring of health status and environmental conditions and threats at home. A very important component of the presented platform, is a central computer station designed to collect and process data in order to evaluate/classify undergoing event alerts. The reconfigurable system consists of a few sensor networks. The basic one is formed by the central station serving as a master, and prescribed devices equipped with appropriate interfaces. Different categories of sensor are developed to measure: heart pulse, temperature, body composition parameters (weight, fat content, etc.), electric heart activity (ECG), and physical activity (accelerometer). Additionally, special sensors/devices can be integrated with the system depending on the type of a chronic illness, e.g. diabetes – glucose concentration, blood hypertension - blood pressure, etc. A set of sensors devoted to monitor essential processes or events include, inter alia: instantaneous water and electrical power consumption, temperature and fire detection, toilet or bath utilization, water and gas leakage. Contact Prof. Jerzy Wtorek, PhD, DSc Department of Biomedical Engineering phone: +48 58 347 13 84, email: [email protected] 18 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Polarization-Sensitive Optical Coherence Tomography with Spectroscopic Analysis Purpose Optical Coherence Tomography (OCT) is one of the most powerful optical methods for non-invasive examination of inhomogeneous objects. Our research interests have been focused on OCT with advanced polarization-sensitive analysis (PS-OCT) and with spectroscopic analysis. It brings unique benefits to OCT measurements, which enable a better characterization of the tested devices and materials. This method is useful, for example in the investigation and quality assessment of complex objects like printed electronic parts or polymer composite materials. Description PS-OCT system PBS 1, 2 – polarization beamsplitters NPBS 1, 2 – non-polarization beamsplitters PP – dispersion compensation prism P – linear polarization plate QWP 1, 2 – quarter-wave plate HWP – half-wave plate OH – optical measurement head PS-OCT measurements examples PS-OCT intensity image of liquid crystal sample; the defected area has been indicated by red contour PS-OCT tomography image of liquid crystal retardation angle Contact Dr. Marcin Strąkowski, PhD Department of Metrology and Optoelectronics Phone: + 48 58 347 13 61 , email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 19 Raman Modular System with Fibre-Optic Probes for Remote Monitoring of CVD Processes Purpose Manufacturing of modern thin-film optical and microelectronic devices in a low-pressure plasma process requires precision and strict control of the process in order to obtain required properties and good quality of the final product. It stimulates the search for effective diagnostic tools. The aim of this presented optoelectronic system is to monitor in-situ growth of thin films in CVD processes. Description Design of Raman system for in-situ monitoring of µPA CVD process: (a) excitation part, (b) acquisition part; 1- coupler, 2 - optical fibre, 3 -excitation probe, 4-6 – objective, 5 – bandpass filter, 7 – window, 8 - growing film, 9 - window, 10-12 - objective, 11 – notch filter, 13 - collecting probe, 14- fibre bundle, 15- adapter; dimensions are given in millimeters. Contact Dr. Marcin Gnyba, PhD Department of Metrology and Optoelectronics phone: +48 58 347 24 82, email: [email protected] 20 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s SART-2: Biometric security system for mobile workstations Purpose Our biometric system increases the security level of mobile workstations by combining face recognition technology with analysis of the user’s behaviour. Mobile devices often serve as data banks containing confidential information, either personal, corporate or governmental. At the same time the risk of unauthorized access to their resources is much higher than in the case of desktop computers. Therefore, additional biometric protection is necessary. Description Our application integrates with Microsoft Windows operating systems and replaces the standard password-based login with an identity verification procedure, performing face identification along with liveness detection. The architecture of the software is modular so that many experimental configurations can be tested within the real-life application. Apart from providing the visual login option, the software monitors a user’s presence and locks the operating system when the computer is left without supervision. Another feature is the possibility to secure access to selected files by means of face recognition. An additional module of our biometric security system detects anomalies in keystroke patterns of a user. As soon as an intruder is detected, the software can lock the workstation immediately or generate a silent alarm, recording the appropriate information in a log file. We are also investigating the possibilities of providing identity verification based on analysis of a user’s habits or patterns discovered in the interaction with graphical interfaces. Our system is fully customizable; thus, we are able to provide the required balance between the supplied security level and the computational complexity of algorithms, which can be adjusted to suit the characteristics of the protected device. Contact Dr. Maciej Smiatacz, PhD Department of Intelligent Interactive Systems phone: +48 58 347 26 89, email: [email protected], http://sart2.eti.pg.gda.pl/ T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 21 Sensor Network for City Traffic Monitoring Purpose The sensor network node, equipped with a low-resolution camera, observes the street and detects moving objects. The number, and the speed of the detected objects, are transmitted using a low-power license-free radio transceiver to another neighbouring node. All the nodes create a self-organized network. The prototype sensor network node has been realized in two versions: FPGA and ASIC. The ASIC version consumes approximately 500mW and it can be powered from a photovoltaic solar panel combined with a single cell Li-Po battery. Description Sensor network: • low power radio transceivers, ISM license-free band; • low power consumption, low hardware and installation costs; • low maintenance costs; • installation on street lamp-poles; • possibility of autonomous power supply. Each sensor network node consists of: • simple camera sensor; • typical transceiver. Advantages: • traffic analysis; • possibility of interaction with traffic lights; • support for emergency vehicles; • detection and localization of traffic jams; • snow information (pictures from cameras); • information for citizens, radio, internet, street information panels; • monitoring of temperature, noise, intensity of the sunlight; • operation in crisis situations. Contact Dr. Marek Wójcikowski, PhD Department of Microelectronic Systems Phone: +48 58 347 19 74, Fax: +48 58 347 23 78, email: [email protected] 22 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Sonar modernization for Navy ships Purpose Upgraded sonar were designed to search for sea mines (MG-89DSP and SHL-200DSP) and for anti-submarine warfare (MG-322DSP, SQS-56PG, SQR-19PG). Description The fast development of electronic technology and digital signal processing methods makes electronic devices, including sonar system ones, obsolete very quickly. But because the ships and mechanical components of sonar are operational for a few decades, it makes perfect sense to modernize these devices by using modern components and signal processing methods, and overhaul antenna and mechanical systems. This approach was used to modernize 17 old sonars for the Polish Navy. In this way, the operational value of sonar was increased, with a financial outlay at around 20% of the cost of new devices. Contact Dr. Jacek Marszal, PhD Department of Marine Electronic Systems phone: +48 58 347 22 30, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 23 “Sound recognition service” – a supercomputer service able to detect, classify and localize threatening acoustic events Purpose The sound processing system was developed as a part of the supercomputer cluster-based “KASKADA” framework intended to facilitate parallel processing of a multiplicity of audio and video streams received from video cameras. The service running on this platform offers the functionality of detecting, classifying, and localizing acoustic events which are symptomatic of threats to the safety of people and property. Description The service is executed in the supercomputing platform KASKADA using a client application, which in turn can run on a PC or mobile platform. The service analyzes live streams from microphones and specialized acoustic sensors. Whenever a threatening event is detected (e.g. a gunshot, scream, breaking glass or explosion), the information is passed to the user via the graphical interface or an sms(text) message. The localization of the event is also calculated and is used to point the camera in the direction of the recognized event automatically. Contact Mr. Andrzej Ciarkowski, MSc Multimedia Systems Department phone: +48 58 347 16 36, email: [email protected] 24 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s System for creating dynamic maps of noise threats employing grid computing Purpose Main purposes of the presented design are: providing detailed information about the noise threats that occur every day in city areas and the prevention of the noise induced hearing loss especially among young people. An experimental system designed for the continuous monitoring of the acoustic climate of urban areas was developed and implemented within the PLGrid Plus project. The assessment of environmental threats is performed based on online data, acquired through a grid of engineered monitoring stations and selected psychoacoustical properties of the human hearing system. Description Dynamic maps of noise threats, including auditory effects which are caused by exposure to noise, are created and updated on the basis of data obtained from a distributed sensors network integrated with the PL-Grid supercomputer infrastructure. Operations are performed employing a dedicated noise prediction model, optimized towards working on a computer cluster. In addition, predicted maps may be adjusted using real noise level measurements. This approach allows for generating more precise maps, but also for verifying mathematical source models, utilized in the process. The work presented resulted in services running on a supercomputer grid, ensuring credible noise distribution results in large areas, displayed on digital maps, that can be dynamically updated. Contact Dr. Józef Kotus, PhD Multimedia Systems Department phone: +48 58 347 29 72, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 25 System for geo-referencing and integration of the spaceborne and airborne real time imagery Purpose The solution is aimed at integrating the spaceborne/airborne, “just captured”, imagery with other types of spatial data in a geographical context. The solution fits any application where the geo-referencing of the observational data and its nearly real time availability is a crucial aspect. Description The main part of the solution are autonomous geo-referencing algorithms which embed the imagery into the geographical context. Algorithms are based on advanced, State of the Art, pattern recognition and computer vision methods. The prototype system involves software and hardware components. The hardware, i.e. platforms and sensors, may be fitted to the user needs. The prototype has been developed, and successfully tested, for different types of imagery and a broad range of spatial resolutions. Particularly, the airborne data was captured from low altitude with the use of a micro UAV platform. The spaceborne imagery was obtained from the polar-orbiting meteorological satellites equipped with the Advanced Very High Resolution Radiometer. Contact Dr. Krzysztof Bruniecki, PhD Department of Geoinformatics phone: +48 58 347 28 85, email: [email protected] 26 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s System supporting speech perception for special educational needs of schoolchildren Purpose Children with special education needs (SEN) represent a large group of pupils (about 4%). Many of these children are not able to understand speech being articulated too fast for them. The engineered system allows for a time scale modification (TSM) of speech signals (i.e. time-expansion) in real-time. As a result, the child could participate in classes with normal-hearing children. Description The system consists of two main parts: the speech-stretching device, and the workstation for performing adequate hearing examination. At the beginning, every pupil has to perform a series of hearing tests (the examination is supervised by the specialist) employing a PC software application. The designed mobile device stretches the speech signal in realtime. It was necessary to design methods allowing the performance of this operation, and to ensure as small as possible the difference in the duration of the input (original speech) and the output (stretched signal). Additionally, dependent on the rate of the input speech, time scale factors are adapted in such a way that fast spoken speech is stretched up more than slow speech. Contact Dr. Piotr Odya, PhD Multimedia Systems Department phone: +48 58 347 23 01, fax: +48 58 347 11 14, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 27 Versatile Real Time Vector-Based Electronic Map Update System for Mobile Devices Purpose This is dedicated, versatile and cheap platform which addresses, for example, vehicle fleet management systems where an interactive approach is required. The platform is small, easy to maintain, and cheap as well. Description The crucial part of the system is the near real time vector map updating subsystem. The mobile part is founded on a static and dynamic layer vector map concept, and the system features a quick and reliable vector map database map merging engine. a) b) Fig. a) System Architecture and b) Front View of the map updating system center The system has already been successfully deployed in a company managing a 200 vehicle (trucks) fleet. Contact Dr. Jerzy Demkowicz, PhD Department of Geoinformatics phone: +48 58 347 25 95 , email: [email protected] 28 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s Video event recognition system with enhanced privacy protection Purpose The system was developed as a part of the supercomputer cluster-based “KASKADA” framework, intended to facilitate the parallel processing of a multiplicity of audio and video streams received from video cameras. The system provides an application for smart surveillance backed up by the supercomputing cluster’s computational power. Its objective is to realize unattended monitoring of urban spaces using a distributed camera infrastructure and advanced video analysis. The system is focused on balancing both privacy and safety in public areas. Description The system was developed using multilayer architecture. In the first layer algorithms are installed, being integrated in the second layer as simple services; subsequently, several simple services are connected in the workflow creating a complex service. A user is allowed to design specific processing scenarios employing implemented algorithms. Instead of video stream transmission, the system generates alerts whenever dangerous events occur. Algorithms developed especially for this system realize object detection, object classifying, people counting, sensitive region detection, congestion, border crossing, restricted area intrusion. The video streams are anonymized, the anonymization algorithm encrypts reversibly the visual content. Contact Mr. Andrzej Ciarkowski, MSc Multimedia Systems Department phone: +48 58 347 16 36, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 29 Virtual Whiteboard Purpose The Virtual Whiteboard is a computer application allowing emulation of an electronic whiteboard using a multimedia projector, a screen, a computer and a simple camera connected to a computer’s USB port. Description The multimedia projector is fixed under the ceiling. A camera can be directly attached to the multimedia projector or placed on a stand at an appropriate distance from the screen, so that it fits in the video frame. The user is situated between the projector and the screen. An application installed on the computer manages the aforementioned components and recognizes dynamic hand gestures, i.e. constituted by motion trajectories, and static gestures, i.e. palm shapes. Apart from the basic functionality of the whiteboard, which is content entering, the system enables the user to interact with objects, e.g. rotation, zooming in/out, cropping and shifting are possible. A course of events during the work with the system can be saved and recreated, preserving time dependencies. The Virtual Whiteboard works with multimedia presentation browsers, providing functionality to browse slides and add notes. Contact Dr. Piotr Odya, PhD Multimedia Systems Department phone: +48 58 347 23 01, fax: +48 58 347 11 14, email: [email protected] 30 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s “Voice Maps” – system supporting independent movement of the blind Purpose This product is a prototype version of a mobile application supporting independent movement of the blind. Its objective is to improve the quality of life of visually impaired people, providing them with navigational assistance in urban areas. Description The mobile application is installed on Android smartphones. Voice recognition, speech synthesis and touch screen enhancements are used as the main user interface methods. The spatial data for the system is collected by trained operators and volunteers. They are stored in a central, remote database in a dedicated format, created to fulfill blind people’s requirements. Movement support is achieved with the use of implemented positioning, navigational and guiding algorithms. Contact Dr. Łukasz Kamiński, PhD Department of Geoinformatics phone: +48 58 347 21 78, email: [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 31 “youarchive.net” – Web-based audio restoration system (Online Sound Restoration System for Digital Library Applications) Purpose Today’s digital restoration and archiving solutions have the following serious limitations: - user needs to have access to the licensed software, - process of sound restoration usually requires a time-demanding procedure, - effective sound restoration needs knowledge, experience and skills from the operator. This system is a web-based audio restoration application which enables ones to automatically explore digital sound libraries faster, restore more easily, and to preserve their content more efficiently. Description The proposed solution has the following features: • no special sound software is needed, • both versions of the signal are archived: original and restored, • restoration could be performed in two modes: • fully automatic noise and distortion reduction • semi-automatic mode of operation • no skills from the user are needed in the fully automatic mode, • limited knowledge of digital signal processing is needed in the semi-automatic mode. The system is dedicated to the removal of typical distortions in music archival recordings, such as: • noise, • impulsive distortions (clicking, based on based on non-linear neural predictor), • parasitic frequency modulation (the “wow” distortion), • signal clipping. Original signal www.youarchive.net Reconstructed signal Contact Prof. Bożena Kostek Audio Acoustics Laboratory phone: +48 58 347 27 17, email: [email protected] 32 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s WeatherSense – numerical weather prediction for Poland Purpose The system provides various information that allow the making of decisions for urban environments based on current or predicted weather conditions. Description WeatherSense is a technology platform that generates and provides high-resolution weather prediction data including various weather parameters such as temperature, humidity, wind, precipitation and others. The WeatherSense Web portal is also integrated with mobile applications designed to serve operational weather predictions to distributed users. Mobile devices deliver tools for weather information display. Currently, three simultaneous technologies are being developed: • Android application, • Windows Phone 7, • iPhone. The utilized numerical model, namely WRF (Weather Research and Forecast), is originally developed by the National Center for Atmospheric Research (NCAR), the Advanced Research WRF (ARW), and the National Center for Environmental Predictions (NCEP). It is run operationally (a 48-hour real-time forecast generated 4 times per day) on the Department of Geoinformatics’ computer cluster. Predictions and weather data generated by the system are available at http://weathersense.pl Contact Dr. Andrzej Chybicki, PhD Department of Geoinformatics phone: +48 58 347 13 26, email. [email protected] T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s 33 Wireless surveillance system module Purpose • prototyping and implementation of hardware accelerated algorithms for HD video and audio compression, • prototyping and implementation of machine vision systems, • prototyping and implementation of parallel data processing systems. Description The presented wireless surveillance system module has been used for development of hardware acceleration for the H.264 video compression standard. In the Xilinx XC6VLX365T FPGA a Microblaze processor with PLB and common memory interface for accelerator modules has been implemented. Linux has been chosen as the operating system for the board. H.264 encoder software has been installed and modified for hardware acceleration. Hardware acceleration has been achieved for: integer transform, inverse integer transform, Hadamard transform, quantization, dequantization, intra-prediction, inter-prediction and CAVLC. ASIC version of the H.264 accelerator (technology 90nm) has also been designed. It can be easily integrated with the development board using the expansion sockets. Contact Dr. Miron Kłosowski, PhD Department of Microelectronic Systems phone: +48 58 347 18 64, email: [email protected] 34 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s The computer laboratories in the new building Inside one of the laboratories in the new building The laboratory of industrial robots in the old building New general purpose computer laboratory opened in 2012 in the new building used for mobile applications development The Clean-room Laboratory of Innovative Materials and Devices with the microwave Plasma CVD System 35 for T h e I n n o v a t i v e F a c uused l t y for f o rgrowth I n n o vofa high t i v e quality Te c h ndiamond o l o g i e films s microelectronic and electrochemical sensors Computer Networks Laboratory in the old building Gdańsk University of Technology Faculty of Electronics, Telecommunications and Informatics Narutowicza 11/12, 80-233 Gdańsk, Poland http://www.eti.pg.gda.pl 36 T h e I n n o v a t i v e F a c u l t y f o r I n n o v a t i v e Te c h n o l o g i e s