Research Interests

Main Research Fields

GeospatialUAVsAudio IntelligenceMachine LearningProgrammable Infrastructures

Automated motion detection from space in sea surveilliance

The Panchromatic Remote-sensing Instrument for Stereo Mapping (PRISM) carried by the Advanced Land-Observing Satellite (ALOS) was designed to generate worldwide topographic data with its high-resolution and stereoscopic observation. PRISM performs along-track (AT) triplet stereo observations using independent forward (FWD), nadir (NDR), and backward (BWD) panchromatic optical line sensors of 2.5m ground resolution in swaths 35 km wide. The FWD and BWD sensors are arranged at an inclination of ±23.8◦ from NDR.

In our research in ADITESS, PRISM images are used under a new perspective, in the security domain for sea surveillance, based on the sequence of the triplet which is acquired in a time interval of 90 sec (45 sec between images). An automated motion detection algorithm is developed allowing the combination of information collected at each instant and therefore the identification of patterns and trajectories of moving objects on the sea; including the extraction of geometric characteristics along with the speed of movement and direction. The developed methodology combines well established image segmentation and morphological operation techniques for the detection of objects. Each object in the scene is represented by unit invariant properties and maintained in a database to allow the generation of trajectories as these arise over time, while the location of moving objects is updated based on the result of neighbourhood calculations.

UAV Platforms Development
UAV Video Analytics

The Web databases, and other digitized information storehouses,, contain a growing volume of audio content. Users and stakeholders want to make the most of this material by searching and indexing the digitized audio content. In the past, companies had to create and manually analyze written transcripts of audio content because using computers to recognize, interpret, and analyze digitized content was difficult. The development of faster microprocessors, large storage capacities and better speech recognition algorithms has made audio mining easier.

In ADITESS, we work towards the development of algorithms that enable users to identify events/words/phrases of interests aiding in increasing efficiency and productivity. Our approach to the solution of problems is the provision of individualized solutions for each particular problem by combining practices and principles from the fields of digital signal processing, audio coding, speech recognition, audio mining and statistical learning.

We assess the effectiveness of our research solutions by utilizing accepted and established standards. We target practical solutions and aim their integration with existing ADITESS products or cots for the production of outcomes that go beyond the proof of concept/technology.

In the hands of experts, machine learning is a powerful tool which can be used for data analytics, pattern recognition, and prediction. ADITESS scientists use state-of-the-art machine learning algorithms and successfully apply them to diverse fields, turning raw data into valuable information, and most importantly into insight and understanding. Big data analytics has become essential practice for modern businesses as it uncovers knowledge which is otherwise concealed in vast collections of data. Data mining and machine learning can be used to discover anomalies, enhance productivity, maximize efficiency, and to forecast opportunities.

A transition from the ‘intelligent design’ approach, which currently rules software engineering to meta-design approaches as well as self-combining software systems emerges from the fact that the “Future Internet” software will be the critical infrastructure on which all other critical infrastructures will depend. The technological developments that are realized are changing the dynamics of the software industry and posing new requirements and opportunities for improving Europe’s competitive position in software.

Our approaches take into account the inability of software engineers to foresee all possible situations that systems, connected to the open physical world, have to face. Therefore, we focus on the design of software components that have the ability to collaborate in an autonomous and decentralized fashion. Furthermore, the deployment of systems and optimisation platforms that take advantage of the emerging programmability of the underlying infrastructure is also of interest for the design of infrastructural-agnostic applications.

Our research relies on the design and development of novel reactive software development paradigms for concurrent and highly distributed applications as well as the design and development of holistic applications’ deployment and management frameworks over programmable infrastructures.