Speakers

Keynote speakers

Take it to the Limit: A story of piezoelectric materials and devices for extreme conditions

Andrew J. Bell
The University of Leeds and Ionix Advanced Technologies Ltd, UK

This presentation will tell the story of how a curiosity-driven, ferroelectric materials research topic developed into a $2 million p.a. business, manufacturing industrial ultrasound transducers for use under extreme conditions. On that journey we discovered some interesting materials science, we learnt how to process materials on the edge of instability, we overcame our academic sensitivities in order to finance a spin-out company and followed some unexpected paths to gain moderate commercial success. The talk aims to provide a contemporary and perhaps idiosyncratic view of both the science and business of piezoelectric materials.

BiFeO3-PbTiO3 is a ferroelectric solid solution with a morphotropic phase boundary (MPB) between the perovskite rhombohedral and tetragonal phases. It has a number of interesting features including the high Curie temperature (635°C at the MPB), an unprecedented spontaneous strain (≈20%) on the tetragonal side of the MPB and the >300 K discontinuity in the antiferromagnetic Néel temperature across the MPB. These features provide a number of exciting opportunities in terms of exploiting the high temperature piezoelectricity and coupling the antiferromagnetism to strain and electric field, whilst the extreme tetragonality provides challenges in materials fabrication. Whilst studying these features the Functional Ceramics Group at the University of Leeds discovered methods to overcome the technical barriers to exploiting the piezoelectric properties and to produce materials capable of “PZT-like” performance up to and beyond 500°C.

As most established piezoelectric materials producers felt that BiFeO3 might be a challenge too far, the University supported the founding of a company, funded by venture capital and private finance, to commercialise the new material. Although there was a market ready for devices spanning the materials’ temperature capability, no potential customers had the capability to build the material into high temperature transducers. Hence, what started as a materials company transformed into a device company, with products in the industrial ultrasound, automotive, aerospace and electronics sectors.

Ground penetrating radar, applications, and design challenges

David Daniels CBE

Visiting Professor, University of Manchester

Ground penetrating radar is a successful development of ultra-wideband short range radar technology and has found application in many areas of the non-destructive examination of visually opaque materials. This presentation will cover its applications, environment, and targets as well as system design, propagation, antenna design and signal processing. GPR provides relative information on targets in the form of images which are often subject to operator interpretation. Many GPRs are essentially uncalibrated instruments providing relative information on target geometry and EM properties. A knowledge of the parameters that affect system performance is key to understanding the information obtained by the GPR. The presentation will reference the application of GPR to landmine detection and provide an update on research at the University of Manchester.

Workshop patrons

ONR Global Overview 

Jay A. Marble

Office of Naval Research Global (London, UK), USA

The Office of Naval Research Global was established just after the Second World War in London, UK.  Today is spans the globe as the international research arm of ONR.  Science Directors (SD) serve as Technology Scouts looking to connect the best of international research to partners in the USA.  Dr. Jay Marble is currently serving as a Science Director for Information, Cyber, and Spectrum Dominance from the London office.  An overview of ONR Global will be given along with highlights from the Electromagnetic Sensor Portfolio.

Infineon / IPCEI presentation

Olga Bulhac

Infineon Technologies, Austria

Semiconductors are crucial when it comes to tackling the environmental, technological and societal challenges of our time and shaping the green transformation. This is why Infineon is committed to actively driving decarbonization and digitalization. However, a commitment like this can only be successfully pursued together. With IPCEI, we want to leverage Europe’s resources and foster collaboration between universities, students, industry experts and start-ups. With our programs, we want to open doors for the bright minds of the future and support them by sharing our knowledge and offering valuable insights into our company. This is what we call our Mission Future! Learn about our IPCEI programs at Infineon Presentation and join our Mission Future.

Speakers

Piezoelectric Anisotropy and Ferroelectric Domain Structure Impact on Sensor and Electronic Element Performance

Marko Budimir

INETEC Institute for Nuclear Technology, Croatia

In modern sensor and electronic technology, the integration of piezoelectric materials offers a multitude of possibilities. These materials exhibit unique electromechanical properties, which, when harnessed correctly, can significantly enhance the performance of various microelectronic devices. This 15-minute talk will delve into the interplay between piezoelectric anisotropy, materials composite structure, and ferroelectric domain configurations, and their crucial influence on the electric properties of sensors and microelectronic elements.

Unlocking the potential of ultrasound technology: revolutionizing healthcare with new excitation methods and advanced semiconductor technology

Harald Gietler

Infineon Technologies, Austria

The advancements in MEMS ultrasound transceivers are opening new possibilities and are making existing applications more accessible. Expensive medical equipment like ultrasound imaging systems can now be replaced by small, portable, and affordable devices. However, this transformative progress comes with its own challenges, such as understanding and modeling MIMO systems as well as the design of MEMS, hardware, and software. New excitation methods in the GHz and THz regime enabled by advanced semiconductor technology improve sensory perception and performance to new heights. This talk wants to provide insights on the current trends and developments in this field, including educational initiatives which aim to prepare young engineers for what is to come; we will shed light on the facets and considerations necessary to adeptly navigate this dynamic and evolving landscape.

Acoustic emissions in plant-hydraulics: from botanist's laboratory towards low-power smart sensor-on-chip for field monitoring of drought 

Dinko Oletić

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

Plant-physiology, agriculture and forestry are searching for new field-instruments to study, quantify, and neutralize effects of ever-spreading seasonal droughts. An interesting non-invasive laboratory method used in early detection of hydraulic failure in drought-stressed vascular plants, is analysis of ultrasonic acoustic emissions (AE) occurring due to drought-induced xylem cavitations. An overview of our ongoing research efforts in transferring AE sensing from plant physiologist’s laboratory test-bench into the field will be given. The first part of the talk describes our design of a wireless energy-autonomous embedded system for in-field AE acquisition and on-board processing. Next, motivated by lowering the power requirements and dimensions, the second part of the talk shows technological research on designing MEMS AE sensors, enabling passive micro-mechanical sorting of AE events by their signal frequency, and integration with low-power electronic detector circuitry.

Deposition of ultra-refractory sputtered layers for high temperature applications: Challenges and some solutions 

Alain  Bosseboeuf
C2N, CNRS, University Paris-Saclay, Palaiseau, France 

Microsensors and other micro/nano devices aimed to be operated at very high temperature (500°C to >1000°C) requires considering the use of ultra-refractory conductive and insulating substrates and thin films for their design and fabrication. This rises various new technological and characterization challenges that must be overcome. This will be illustrated in the case of ZrC substrates and Zr-based oxide, nitride and carbide films deposited by sputtering, reactive sputtering and co-sputtering. It will be notably shown that elaboration of stoechiometric films with suitable mechanical, electrical and properties is hard to achieve for many reasons related to film composition measurements, to oxygen gettering effect by zirconium, to residual and thermomechanical stress, to abnormal film composition, etc. Solutions investigated to characterize and solve these various issues will be discussed.

A low-power data logger for perishable goods transport monitoring

Marko Gazivoda

Intis, Croatia

We present Nutrico ET, a single-channel low-power temperature and vibration logger with an average power consumption of around 25 μW. This device is a part of an innovative logistic cold chain system used to supply the end users with fresh and highly nutritious food products at their workplace. Such a cold chain is by law required to provide data on perishable goods temperature during transport and storage, which is achieved using this device. The low power consumption and small log sizes allow the devices to function uninterrupted for long periods of time, while its small size and mass, along with its waterproof casing make it practical for use with any type of perishable goods packaging.

Road monitor: A low-power wireless sensor node for long-term monitoring of road structural integrity

Fran Penić

X-Logic, Croatia 

At X-Logic, we have developed the Road Monitor, a low-power wireless sensor node for monitoring the structural integrity of road surfaces. Using two accelerometers buried at different depths, the Road Monitor is able to monitor the deflection of both the surface layer and the subgrade of the road caused by passing vehicles, sending the collected data to a server via NB-IoT. In contrast with the methods currently used for road deflection measurement, the sensor node is fully embedded into the road surface, enabling long-term continuous monitoring without impeding traffic. By continuously monitoring road conditions, road maintenance can be realized before visible damage occurs, leading to increased road safety and lower repair costs. 

Magnetic field measurements: from RTD-Fluxgate, through the IR3MA Project, to emerging solutions

Carlo Trigona

University of Catania, Department of Electrical, Electronic and Computer Engineering, Italy

Magnetic field measurements have garnered significant interest across various application fields, including environmental monitoring, localization, security, and medical applications. Within the realm of medical applications, a particularly compelling focus lies in utilizing these measurements for the diagnosis of neurodegenerative diseases. In this context, the application of RTD-Fluxgate technology to measure the concentration of iron in the human brain holds great promise.It is important to emphasize that this approach enables the estimation of iron concentration, which in turn facilitates the diagnosis of conditions like Parkinson's disease, as well as the monitoring of degeneration in patients. This method not only holds potential for early disease detection but also offers the ability to track iron accumulation, providing crucial insights into disease progression. Furthermore, emerging solutions with a stronger focus on environmental sustainability, through the adoption of living sensors, will be presented for measuring physical quantities of interest, encompassing static and quasi-static magnetic fields. These innovative solutions play a crucial role in advancing measurement techniques, thereby improving their precision and reliability across various applications.

Eddy Current Testing Sensor for HSS Drill Tool: Experiment & Simulation 

Athanasios Kyrgiazoglou

University of Western Macedonia, Mechanical Engineering Department, Greece

Machining technology performs a major role in modern manufacturing industries. A key bottleneck associated with any type of machining is the accelerated tool wear and the resulted reduced tool life. Most of the drill-bits made by conductive materials and as a result the eddy current testing method (ECT) can be applicable. Depending on the mechanical and physical properties of the workpiece, different drill-bits are used with the greatest differences between them being their core material and their coating. The ECT measurements produce very informative results. The signal’s amplitude and the phase lag give a lot of information about tools edge condition. Moreover, the good agreement between experimental and simulation results extract conclusions that allows the use of theoretical research for new sensors development and in addition, to understand better the influence of the tool wear as well the method limitations in the drill-bit tools ECT inspection.

Identification of non-ferrous metals using magnetic induction spectroscopy on an industrial conveyor system

Michael Mallaburn

University of Manchester, UK

A prototype system has been built and tested to enable identification of different classes of non-ferrous metals using magnetic induction spectroscopy on an industrial conveyor system. Results from a non-moving static test system has shown recovery and purity rates typically >80% for various types of waste streams. In contrast, preliminary tests on a moving system have shown somewhat degraded results with the best class of metals (stainless steel) having purity and recovery of ~60%. This presentation covers some of the recent improvements aimed at increasing the performance of the conveyor-mounted system whilst looking forward to another trial to test the capabilities. Also shown are ideas for where to go next with improving the discrimination capabilities for non-ferrous metals.

Efficient and Accurate Computation of the Magnetic Polarizability Tensor (MPT) Characterisation of Realistic Metallic Targets.

James Elgy

Keele University, Staffordshire, UK

In this talk, we will provide an overview of our computational approach to accurately and efficiently compute the Magnetic Polarizability Tensor (MPT) spectral signature characterisation of different metallic objects, for application in metal detection inverse problems. This will include combining a hp-finite element methodology with a proper orthogonal decomposition (POD) reduced order modelling technique for fast computation of spectral signatures and the handling of highly conducting and magnetic objects with thin skin depths using prismatic boundary layer approaches. Realistic examples will be included as well as a brief overview of our open-source MPT-Calculator software.

Combining machine vision and magnetic induction sensors to improve classification of scrap metal on a conveyor belt 

Kane Williams

University of Manchester, UK

Magnetic induction sensors have produced promising results when classifying scrap non-ferrous metals of different conductivity. However, magnetic induction struggles to separate metals of similar conductivity. A camera used to extract the colour of the metal pieces has been shown to improve the classification of metals with similar conductivity due to the colour contrast. The camera and induction sensors must communicate in real-time and have all the measurements sent to the STM32 microcontroller, where the machine-learning model is processed. This process continuously occurs while the conveyor moves 2M/s with different metals of shape, size and surface contamination.

Piezoelectric energy harvesting for autonomous sensor networks 

Saša Zelenika

University of Rijeka, Croatia

An overview of the fields of application of energy harvesting technologies, with the emphasis on its importance and integration into sensor nodes, will be provided in this talk. An in-depth presentation of the possibilities of using kinetic energy harvesting, in the form of piezoelectric devices, will follow. Aspects related to the respective solutions implemented in structural health monitoring, river pollution monitoring, tire pressure monitoring and biomedical application will be given. Issues concerning the modelling of the behaviour of the considered class of devices, of the optimisation of their performances and of the experimental assessment of the resulting power levels will also be provided.

Exploring the potential of emerging applications of mmWave polymer waveguides in communication and sensor technology

Siegfried Krainer

Infineon Technologies Austria AG

Thanks to the remarkable advancements in Automotive Radar systems and technologies, the development of mmWave Excitation and modulation has witnessed great progress, pushing towards higher frequencies in the THz regime.

Making use of SiGe Technologies, circuits are able to achieve noticeable output power, efficiency and noise characteristics. By harnessing modulated mmWaves as information carriers, intriguing possibilities arise, bridging the gap between fiber optics and wireline communication systems. In this talk, we will present the outcomes of communication demonstrators which leverage Polymer Fiber Technology in 60-80 Ghz and 150 Ghz; these findings carry significant implications for Datacenters and other emerging tools. Furthermore, considering the high interaction of mmWaves with biological matter, we will provide an overview of sensor-related opportunities for biological applications.

Development of two-dimensional fuel cell

Krešimir Dželalija

University of Split, Croatia

Our research focuses on the development of a compact two-dimensional micro-hydrogen fuel cell using SU8 photoresist. Our home made proton exchange membrane hydrogen fuel cell (PEMFC) has hydrogen and oxygen channels on the same side of the electrolyte membrane. The channels (200 μm in width) with a total volume of about 2.4 mm3 are separated by a 120 micrometer thick wall and covered with a membrane. Through experimental investigations, we have obtained promising results for the performance of the fuel cell. These results demonstrate the effectiveness of our two-dimensional micro-hydrogen fuel cell. The compact integration of channels on one side of the electrolyte membrane enables the future development of 2D fuel cell stacks. This research contributes to the advancement of sustainable hydrogen-based energy conversion technologies, with potential application in low power microelectronics.

Research and Design Challenges in 3D Fringe Projection Profilometry 

Tomislav Petković

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

Fringe projection profilometry (FPP) is a popular technique for non-contact measurement of the 3D surface profile of an object. In FPP one or more fringe patterns are projected onto the object and the 3D surface profile is then reconstructed from the observed deformations of the projected fringe pattern. Important questions of interest in FPP scanner design among others are: What is the practical optical setup or how to spatially arrange cameras and projectors? How to select fringe parameters to improve both accuracy and robustness? How to tackle unwanted interferences including ambient light and multiple reflections? This talk will give a short overview of common approaches to resolve these challenges in FPP scanner design and will discuss the related open research questions. An emphasis will also be placed on the underwater FPP 3D scanning which is in the current research focus at the University of Zagreb. 

PhD Presentations

Wet etching of ZnO piezoelectric material for MEMS vibrational sensors 

Zarina Nazyrova 

C2N, CNRS, University Paris-Saclay, Palaiseau, France 

Piezoelectric materials are widely used in sensing applications. For micromachined sensors, wet etching has the advantage of being faster and more setective than dry etching techniques. However, the crystalline structure of ZnO leads to anisotropic chemical etching with different lateral and vertical etch rates that makes more complicate its patterning by micromachining. In our talk, we will make a review of our results on ZnO etching  by various wet etchants with the specific target of defining a MEMS resonator piezoelectric layer process step. 

Pocket Therapist: Combining smart wearable sensors with real-time machine learning to provide emotion detection and regulation 

Charis Whyte 

University of Manchester, UK   

Combining physiological data collected via smart wearables, with machine learning, the pocket therapist project aims to classify emotional states in real-time, with an end goal to assist Autistic individuals regulate their emotions. This talk will cover some context of the existing technology as well as the target groups, before diving into the technology being designed and used in the project thus far. Pairing a smartwatch with a designed android app; these signals are collected and analysed to calculate the chance of extreme emotions occurring, for which then prompt the user for feedback to build a user profile. This is the groundwork preparing for long term study with autistic adults to take place towards the end of 2023. 

Towards more accurate detection of human falls

Krunoslav Jurčić, Ratko Magjarević

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

We present a research conducted on different biological data types using various techniques of data analysis in order to create the best performing algorithm for recognition of human activities, with special attention being given to detection of unwanted hazardous events, more specifically human falls. The motivation for the development of physical activity recognition algorithm includes helping the elderly population in case of unexpected activities. In this case, real time fall detection could help in alerting the caregivers of unwanted physical activity, and therefore providing the elderly living alone with the necessary help more quickly and efficiently. 

Comparison of Simulated and Measured Results of Non-contact Capacitive Electrodes for Biomedical Applications 

Luka Klaić, Mario Cifrek  

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

In biomedical applications, signal acquisition heavily depends on the characteristics of the skin-electrode interface. Unlike wet and gel surface electrodes, dry electrodes do not require the presence of electrolytic gel. More specifically, capacitive dry electrodes are contactless, i.e., they can be applied over clothing, which makes their application easier, faster, less intrusive, and less painful. When such electrodes are considered, the analysis of the geometry and electronic design is one of the first steps in the process of increasing the immunity against environmental noise and motion artifacts. The goal of this research was to develop an electromagnetic behavioral simulation model of the four-layered capacitive electrode and explore its reliability via a comparison of the simulated and measured results when geometrical and fabric parameters are changed.The results have confirmed the possibility of using the developed model for qualitative system analysis. The drawn conclusions could be used as guidelines for more efficient capacitive electrode prototype design in order to accelerate development, as well as save time and resources. Future work will focus on the question of precision and repeatability of the obtained results. 

Measurement of Implantable Capacitive Intrabody Communication Channel Properties on a Liquid Phantom

Matija Roglić, Željka Lučev Vasić  

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

In an everyday scenario, medical devices that are used for biomonitoring must be energy efficient but also provide security of private data that they track. Capacitive intrabody communication is a method that provides a safe and energy efficient way of data transmission between devices on and inside human body, but is susceptible to environmental changes. In this work, we are using a liquid phantom representation of human muscle tissue to measure and characterize the communication channel between an in-body and an out-body devices of an implantable capacitive intrabody communication system. 

Thermal conductivity measurement techniques in anisotropic thin film materials

Anton Kabaši

University of Split, Croatia

Goal of our research is development of a proton exchange membrane (PEM) with increased lateral thermal conductivity to improve hydrogen fuel cell efficiency and longevity. PEMs in polymer electrolyte fuel cells belong to a class of materials known as ionomers. These materials exhibit anisotropic thermal conductivity which depends highly on chemical composition, degree of hydration, size of water domains surrounded by polymer chains whose geometry is influenced by the stress and strain induced in the membrane during operation by local overheating. Frequency domain measurement techniques (3ω method) with arbitrary sensor geometries enable measurement of the general thermal conductivity tensor. Micro heater sensors are fabricated by photolithography onto a measured material. AC current is passed through the micro heater with frequency ω, which induces local heating of the material as a function of 2ω. Resistance depends linearly on the temperature change. Voltage differential across the metallic heater is measured. The third harmonic of the voltage differential contains information on the thermal properties of the material. 

Design of composite resistors in 180-nm CMOS technology for RC oscillators and voltage references

Ivana Franković, Franjo Mikić, Josip Mikulić, Adrijan Barić   

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

The temperature coefficient of a resistance can be adjusted by using composite topologies which combine two or more resistors of opposite nominal temperature coefficients. Five different topologies are  analyzed and compared to determine which topology has the best performance with regards to the process variations and required area. Two of the topologies are designed to achieve zero temperature  coefficient and implemented within a 1-MHz RC oscillator. These two topologies are additionally modified to have trimmable temperature coefficient and implemented within a voltage reference. 

Characterisation of various integrated silicon stress sensors

Leo Gočan, Adrijan Barić    

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

In order to correctly measure the stress in silicon chips, the piezoresistive coefficients of the silicon used by the specific technology need to be extracted. For that purpose, various transistor and resistor integrated silicon stress sensors can be used. By arranging the sensors of different sizes and types in a matrix, their performance with respect to their properties can be extracted alongside the piezoresistive coefficients of silicon. 

The method for the extraction of stress-related characteristics of ICs caused by the packaging 

Andro Žamboki, Adrijan Barić    

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

Integrated circuit packaging generates a static stress over the silicon devices which can have detrimental impacts on integrated circuits (ICs), especially biasing sources and oscillators. This presentation will outline the method we have developed for the extraction of the stress responses of the ICs. The results from this method can be used to calibrate and enhance stress-dependent models so that the stress response of circuits can be simulated. With this, active stress-compensating circuits can be designed so that stress-related problems can be reduced before manufacturing. 

Landmine Identification From Pulse Induction Metal Detector Data Using Machine Learning

Marko Šimić, Davorin Ambruš

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

Significant presence of metallic clutter in minefields results in a high false alarm rate of conventional metal detectors due to their inability to discriminate between metallic parts of a landmine and non-hazardous clutter. The presentation describes an electromagnetic induction (EMI)-based system for identification of small hidden metallic objects. A 1D-convolutional neural network is employed to infer an object class from time-domain magnetic polarizability tensor features. The proposed approach is experimentally validated under laboratory conditions on dataset including two types of landmines and five types of metallic clutter. The model trained on simulations and tested on measurements achieves 98% accuracy (with zero false negatives) for both multi-class and binary “threat or nonthreat” classification problems. 

High-frequency Electromagnetic Induction Sensing of Soil Electrical Conductivity and Dielectric Permittivity

Dorijan Špikić, Darko Vasić

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

Electromagnetic induction modality for soil sensing applications in the high-frequency range from 3 MHz to 30 MHz can result in higher sensitivity to electrical conductivity and dielectric permittivity. We present a high-frequency electromagnetic induction (HFEMI) sensor, made with shielded printed circuit board coils, in the entire high-frequency range for a typical range of values of soil electrical conductivity and dielectric permittivity. The HFEMI sensor is analyzed experimentally in the laboratory and the field, and the results were verified using finite element analysis (FEM). In the laboratory part of the study, we measured the sensor response above a container filled with deionized water whose conductivity and permittivity were controlled by adding salt and sucrose. The results are compared with the FEM analysis, and they indicate the possibility of the simultaneous determination of the soil electrical conductivity and dielectric permittivity. The model of a double-layered medium and the concept of an effective medium were used in the second laboratory experiment to obtain lower dielectric permittivity values typical for soil. In the field experiment, the sensor response was measured above the soil irrigated with deionized and saline water. The results are in agreement with the laboratory experiments. 

Non-contact Capacitive Electrode Systems for Biomedical Applications

Antonio Stanešić, Mario Cifrek

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

Non-contact capacitive electrodes have gained importance recently due to their minimally invasive interface which requires no direct skin contact, showing great potential for the contexts of wearables and non-invasive health monitoring. Such electrodes can be embedded into clothing, bedding, seating etc. These systems also present many challenges, such as ensuring the integrity of the signal and sufficient signal-to-noise ratio, minimising motion artifacts, capacitive artifacts, and other sources of noise. The goal of this research was to design and develop pairs of the four-layered capacitive electrode and investigate their performance and stability while exploring the dielectric and geometrical parameters of the system. The results have confirmed the possibility of using the developed electrodes for biopotential measurements, while indicating possible optimal geometrical parameter intervals and limitations of the design. Some completely new control parameters and variables were noted, which will greatly impact future design and research. This research enables design of advanced methods for artifact and interference elimination, as well as further research into stability and temporal integrity of the system. 

Customized sensing platforms for dielectric cure monitoring 

Dailys Arronde Pérez 

University of Klagenfurt, Sensors and Actuators Group (SAT), Austria

Cure monitoring is a means of tracking the real-time changes in physical state or chemical reaction that occur during the curing process of resins. Dielectric Analysis (DEA) is the more effective method to measure composite cure state in real time under actual conditions, which can sometimes involve harsh processing environments as ovens, presses, autoclaves, etc. DEA, also referred as dielectric cure monitoring measures the variations in the dielectric properties of thermosets or polymers, relating the electrical quantities to physical properties such as viscosity, modulus and cure state. In this work, dielectric cure monitoring has been performed using self-developed sensing platforms using different signal processing frameworks. The different sensing principles and signal processing devices used to monitor the curing process, as well as the results obtained during their experimental evaluation will be presented.

Calibration of a Structured Light Imaging System in Two-Layer Flat Refractive Geometry for Underwater Imaging

Domagoj Zoraja, Tomislav Petković

University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia

The development of a robust 3D imaging system for underwater applications is a crucial process in underwater imaging where the physical properties of the underwater environment make the implementation of such systems challenging. Calibration is an essential step in the application of such imaging systems and is performed to acquire the parameters of the image formation model and to enable 3D reconstruction. We present a novel calibration method for an underwater 3D imaging system comprising a pair of cameras, of a projector, and of a single glass interface that is shared between cameras and projector(s). The proposed calibration uses a numerical optimization of a 3D cost function to determine all system parameters, thus avoiding the minimization of re-projection errors which require numerically solving a 12th order polynomial equation multiple times for each observed point. The proposed calibration was experimentally evaluated on four different glass interfaces, wherein several quantitative results were reported, including the re-projection error. The achieved mean absolute errors for the reconstruction of a flat surface were 1.38 mm for normal glass interfaces and 2.82 mm for the laminated glass interface, which is more than sufficient for application.