Servicio de Información Comunitario sobre Investigación y Desarrollo - CORDIS

Final Activity Report Summary - DENIDIA (Development of Excellence in Non-Invasive Diagnostic Systems for Industrials and Scientific Applications)

With 7 recruited fellows and 8 training periods in 6 different top-class partner institutions for a total of 122 research-months, the DENIDIA project has successfully achieved its original goal of increasing excellence in tomography technology and related applications by covering a large spectrum of R&T activities in hardware design and development, image reconstruction, image processing and analysis. All these activities have participated to the important mass of knowledge that has been transferred to CED of TUL in multi-disciplinary fields of science such as electrical and electronic engineering, multi-phase flow and chemical engineering, materials science, nuclear science, instrumentation and technology, and of course, computer science.

All the activities were organised within 3 main actions:
Action 1: 3D ECT - Sensor design and IT environment,
Action 2: Spatial resolution improvement of ECT - new concepts and design.
Action 3: Software development for advanced 3D visualisation, 3D image processing and 3D data analysis.

A major result has been the design of a new multi-purpose Electrical Tomography acquisition unit with 32 channels, called DECART. Originally designed for 3D capacitance measurements (action 1), it has been updated for parallel measurement of capacitance and resistance between pairs of electrodes in the multi-modality sensor which also incorporates gamma-ray measurement. This unique sensor, initially designed and simulated at TUL by prof. Hammer and Dr Mosorov, then developed by Dr Nowakowski in collaboration with UiB, is able to determine component fractions in stratified 3-phase flow (i.e. oil/salt water and air) (action 2).

DECART has also the great advantage, since developed at TUL, to allow synchronisation of the capacitance measurements with rotation of the rotatable ECT sensor developed by Dr Liu (action 2). The results from this innovation have shown that image accuracy of static object is increased thanks to the increased capacitance measurements obtained during rotation, while some limitations linked to the acquisition speed and low SNR have reduced the ability of the sensor to better quantify real 2-phase flow regimes. All images from the above-mentioned sensors are visualised on line thanks to the GUI software developed at TUL, that uses standard image reconstruction algorithms, but also the one developed by Dr Banasiak at UoBa (action1).

These algorithms, designed for 3D ECT sensors and based on FEA, integrates the electrode layout obtained from ElecNet CAD software to better estimate shape contours than state-of-the-art approaches. Compared to action 1 and action 2 that have been relatively close in terms of research subjects dealing essentially with ECT-based research, action 3 has been a relative separate action. A link with action 1 still exists with the comparison between X-ray tomography data and ECT measurements during silo discharging. However, most of the tasks have concerned the development of image processing algorithms, as well as data analysis, related to materials science application based on X-ray tomography measurements.

More specifically, a so-called hole filling algorithm developed by Dr Janaszewski in collaboration with ESIEE was very useful to fill holes in 3D materials microstructure, helping image analysis in different studies carried out in collaboration with UoM. Another aspect of the work carried out by Dr Kornev in collaboration with ESIEE has concerned surface reconstruction of fractured object to separate surface crack from surrounding. Another aspect of the action concerns the development of a 3D stereoscopic vision system by Dr Tan. This equipment, preferentially developed for X-ray tomography images, is also of great interest to visualise results from other modalities such as 3D ECT.

Reported by

Síganos en: RSS Facebook Twitter YouTube Gestionado por la Oficina de Publicaciones de la UE Arriba