ez itt az index

Computer graphics and machine vision

Activity of the research group:

The Computer Graphics group focuses on real-time GPU-based photorealistic image synthesis, off-line CPU-based global illumination, non-photorealistic/artistic imaging simulation, medical/physical/engineering systems simulation and visualization, image processing, machine vision, and tomographic reconstruction.

Recent results:

• Dynamic PET reconstruction system algorithms and GPU implementation
• GPU-based sampling and reconstruction
• Non-Euclidean geometries in games and virtual reality systems
• Real-time algorithmic photography: fog, rain, etc. Elimination
• Deep learning applications in image processing and graphics
• Retinal transfer function determination (collaboration with Botond Roska’s group in Basel)

Former students have gone on to professional management positions at Intel, Google, Amazon, Weta Digital, Industrial Light&Magic, Disney, DreamWorks, etc.Attila Áfra, a former PhD student and current Intel researcher, won this year’s Technical Oscar, the highest award for practical application of computer graphics.

Special infrastructure:

GPU server • Reconstruction server • GPU workstation

Recent projects:

VKSZ • VEKOP • VKE • OTKA

International relations:

Universidad de Girona • Cseh Műszaki Egyetem • Auckland University of Tech. • Pompeu Fabra University • NMI Reutlingen • Griffith University • Universitat Wien

Industrial partners:

Mediso

Radar and Remote Sensing

Activity of the research group:

Microwave active and passive remote sensing

• decision and estimation theory
• classification
• antenna systems
• RCS fluctuation
• micro-Doppler
• radar modulation
• passive radar
• holographic radar
• CUAV radar
• WAMLAT
• RF, KF, digital KF, SP circuits

Smart antenna systems

• digital beamforming, electronically scanning antenna
• Radio direction finding techniques
• Interference and jammer filtering techniques
• multibeam antennas
• Antennas and antenna systems realisation

Sensor fusion

• sensor fusion theory
• CUAV applications
• object protection and border surveillance applications

Small satellites and ground stations

• Automated and remote controlled satellite control and receiving stations
• educational small satellite: power supply system, on-board computer, radio communication system, experiment as payload
• satellite system integration
• small satellite operation

Recent results:

• Holographic radar demonstrator for drone detection
• Passive radar demonstrator using fused FM and DVB-T illuminator
• Micro-Doppler demonstrator
• radio direction finding demonstrator
• Radar tester demonstrator
• Smart antenna system demonstrators
• CUAV system
• wave propagation studies in densely vegetated areas
• Investigation of broadcast transmitters’ radiation intensity in the altitude range 0-kb.30km
• First four Hungarian satellites at the Technical University of Technology: MaSat-1, Smog-P, ATL-1, Smog-1
• BME full remote satellite ground station

Special infrastructure:

Rohde & Schwarz ZNA Vector Network Analyzer 10 MHz to 43.5 GHz • Rhode&Schwarz Vector Signal generator SMBV100B • Rhode&Schwarz Signal Analyzer FSV • Agilent Vector Signal generator E4438C • Agilent S Parameter Network Analyzer 8753ES • Rhode&Schwarz Spectrum Analyzer FSH3 • Oscilloscopes

Recent projects:

Piaci KFI

Industrial partners:

MH Institute for Modernisation • NKE • ATOMKI • Hungarocontrol Zrt. • WIGNER • BHE Kft. • metALCOM ZRt. • Pro Patria Electronics Kft. • Tungsram Operations Kft. • MVM Paks II. Zrt.

Microwave and optical telecommunication systems

Activity of the research group:

Wave propagation, systems engineering and physical layer aspects of bulk antenna techniques. Investigation of massive MIMO systems. Cognitive radio, modulation techniques for opportunistic spectrum usage, signal processing issues, software radio applications. Validation of electromagnetic field computation models for radio devices. Modulation, coding and synchronisation issues in satellite communications. DRM30 and UHF DVB coverage optimisation.

Recent results:

• Development of the DSP part of the ground station for the Alphasat satellite reception experiments
• Development of telemetry for SMOG-P/SMOG-1/ATL-1 satellites, reception software
• DARPA Spectrum Collaboration Challenge results
• Participation in the development of the IEEE/IEC 62704 standard
• Optimisation of the Hungarian DVB-T2 network (with Antenna Hungária)

Special infrastructure:

AWR Microwave Office • DRM Content Server • VNA 40GHz / 4GHz (HP) • VSA 3.5GHz, 7GHz (R&S FSIQ) • EMI transducer (R&S ESCS) • Spectrum analysers up to 7 GHz • SDRs (USRP X310 2 db, PlutoSDR ~10 pcs, USRP2 4 pcs, USRP 2 pcs) • Communication analysers (R&S CMU, EFA, DVG, DVMD, …) • Broadcasting transmitters (DVB-T, FM, DAB) • Oscilloscopes, signal generators • DWDM set • Basic optical instruments

Recent projects:

ESA, Alphasat • EFOP

International relations:

Joanneum Research, Ausztria

Industrial partners:

Ericsson Hungary • SZOMEL Kft.

Digital signal processing

Activity of the research group:

Our aim is the development and research of signal processing methods, and, besides, the utilization of high-level knowledge in measurement technology, electronics. The speciality of our laboratory is the application of digital signal processing to acoustic problems. Active noise control is one of the first research works that has been shortly accompanied by sound synthesis and audio signal processing. The research is supported by the equipment of the laboratory comprising more and more vibroacoustic sensors and other instruments.

Another important research field is the frequency domain identification and some related topics, like parameter estimation, linear and nonlinear system identification. Research and development covers the field of baseband complex signal processing in wireless systems, especially in the topics of 5G modulation, channel estimation and parameter estimation.

Recent results:

• Audio filter design: Investigation of finite wordlength effects of fixed-pole parallel filters, their extension for MIMO systems. New methods for conversion of IIR filters to second order sections based on direct or serial form.
• Development of a noise canceling office chair, investigation of active noise control systems with multiple reference signals.
• Adaptive systems: Exhaustive investigation of the observer-based adaptive Fourier analyzer.
• Distributed systems: New method for spectral estimation in the case of data loss, frequency domain identification of data loss models.

Special infrastructure:

Agilent E5061B network analyzer • Analog Devices signal processing cards, development environments • Brüel & Kjaer vibroacustical measurement equipment

Recent projects:

Biztonságtudományi és technológiai kompetencia központ • EFOP

International relations:

Karlsruhe Institute of Technology • Aalto University • KU Leuven

Industrial partners:

Ericsson • ProDSP Zrt. • Thyssenkrupp Components Kft. • Knorr-Bremse • Óbuda University

Acoustics

Activity of the research group:

In the Laboratory of Acoustics and Studio Technologies we are mainly concerned with the measurement, modeling, and prediction of sound and vibration propagation in many areas of application. The research profile of the laboratory includes vibroacoustics, numerical methods, virtual acoustics, psychoacoustics, musical acoustics, and aeroacoustcs. In these areas we have achieved both theoretical and practical results.

Recent results:

• A generalized theory of sound field synthesis has been elaborated, that includes previous, isolated theoretical developments and allows for extending them. The PhD dissertation of Gergely Firtha and numerous international journal papers resulted from this topic. (Reproduction of a realistic sound field in an extended area. Applications: entertrainment industry, active noise cancellation.)
• Development of a drone detection system based on acoustical beamforming, which is capable of the automatic detection of the angle of arrival of UAVs. The sensor and its framework is able to cooperate with other sensors based on different detection principles. (A demo has been prepared, the drone is detected using radar, optical sensors, acoustical beamforming, and radio signal measurements simultaneously.)
• Development of a Virtual Community Tool that is capable of predicting the effects of the changes in airport traffic on the lives of inhabitants. (ANIMA H2020 project)

Special infrastructure:

Accelerometers, microphones, impact hammers, calibration equipment • 48-channel microphone array with data collector • Measurement room with half open sound space

Recent projects:

H2020 ANIMA • Bolyai • ÚNKP

International relations:

TU Rostock • Fraunhofer IBP

Industrial partners:

Robert Bosch Kft.

3D geometric modelling

Activity of the research group:

Digital representations of complex 3D geometric objects (point sets, triangular meshes, curves and curve meshes, parametric and implicit surfaces, solids), related mathematical algorithms and applications.

Recent results:

• Multi-sided surfaces: We have developed several surface representations over the years, both transfinite interpolation surfaces and control point based patches. Our novel, generalized Bézier and B-spline surfaces can interpolate arbitrary boundary ribbons and permit interior hole loops, as well.
• Reverse engineering: In the majority of engineering applications, it is crucial that the reconstructed models satisfy various geometric constraints. The primary surfaces must obey various rules, such as being orthogonal, parallel, tangential, symmetric, concentric, and so on. If we approximate the segmented regions individually, one by one, we may obtain inaccurate surfaces and poor CAD models. The goal of this research is to perfect CAD models created from measured data. We introduced techniques to automatically detect likely engineering constraints, and enforce these by performing constrained fitting.
• Parameterization & Surface fitting: During the reverse engineering of CAD models, free-form regions need to be approximated with trimmed tensor-product B-splines. There are many methods available for surface fitting, but all of them require an initial parameterization (flattening) of the data points, which can have a substantial effect on the quality of the final surface. While mesh parameterization has a rich literature, the specific requirements of trimmed fitting are quite different from those of texture mapping or quad meshing. Motivated by this, we worked on mesh parameterization methods that enable fitting high-quality trimmed surfaces with minimal user intervention.
• Proximity curves & surfaces: We introduced two different representations to define proximity curves. The representations satisfy many important requirements that are useful in CAD, including positivity, convex combination, natural knot insertion and so on.
• Implicit surfaces: Implicit surfaces provide an interesting alternative to parametric patches. Many operations (e.g. inside-outside testing, Boolean operations, ray-tracing, connecting to regular implicit surfaces) are more convenient using them. We have explored the I-patch representation for various applications including polyhedral design and data approximation.

Special infrastructure:

FormLabs 3D printer

Recent projects:

OTKA

International relations:

Shizuoka University

Industrial partners:

Boulder Graphics LLC • Shapr3D • FormLabs