Projects


Research projects the Visual Computing Group has participated in.

CameSense — Advanced Use of a Swarm of Unmanned Aerial Vehicles Leveraging Augmented/Virtual Reality and Artificial Intelligence Technology for Sustainable Camelina Cultivation in Northern Greece (2025–2026) — a precision-agriculture project building a UAV swarm platform for monitoring camelina sativa crops. Through a mission-planning interface, operators define flight parameters and dispatch drones to scan fields for weed detection and crop health assessment. The Visual Computing Group (DUTH-VCG) participates as one of two Democritus University of Thrace laboratories in the consortium, alongside DUTH-ConvCAO, BIOS AGROSYSTEMS SA and INTELLIA.

Visit project ↗

ASPIDA — Improvement of the Quality of Life and Activity for the Elderly — an integrated system combining wearable electronics, image processing, robotics and smart-home IoT to support independent living for elderly users. It targets real-time health and fall-risk monitoring, personalized physical rehabilitation, and robotic assistance for emergency response, through a simple interface aimed at elderly users, caregivers and healthcare professionals.

The project is implemented under the Support for Regional Excellence Program (Grant MIS 5047294), in part under the Competitiveness, Entrepreneurship and Innovation Program (NSRF 2014-2020), and is co-funded by Greece and the European Union (European Regional Development Fund).

Visit project ↗

Hellenic Autonomous Vehicle — a Democritus University of Thrace project to research and develop the first Greek autonomous vehicle (SAE Level 3+), for both research and commercial purposes, integrating a real-time perception, planning and control stack (HD maps, multi-sensor systems and on-board computing). The Visual Computing Group is among the participating DUTH groups.

The project is implemented under the Operational Program Competitiveness, Entrepreneurship and Innovation, under the call RESEARCH-CREATE-INNOVATE (project MIS: 5069164) and is co-funded from the European Union and National Funds.

Visit project ↗

μDoc.tS — a platform for the transcription of historical handwritten manuscripts, developed as a Greek “RESEARCH–CREATE–INNOVATE” national project (project code T1EDK-01939) with the participation of the Visual Computing Group.

Read more →

Content-Based Image Retrieval for Computer-Aided Diagnosis — work by the Visual Computing Group (L. Tsochatzidis, I. Pratikakis) incorporating content-based image retrieval into the computer-aided diagnosis of mammographic masses: a two-stage, margin-specific, SVM-based CBIR scheme using a dedicated texture descriptor (mHOG), evaluated on the DDSM mammography database.

Visit project ↗

PREdictive digitization, reStoration and degradatIOn assessment of cultUral heritage objectS — an EU FP7-ICT project (2013–2016, coordinated by NTNU, Norway) that developed predictive digitisation technologies for cultural heritage: speeding up and simplifying 3D digitisation, quantifying the degradation of stone monuments, and reconstructing objects from their constituent fragments.

Visit project ↗

Recognition and Enrichment of Archival Documents — an EU Horizon 2020 project (2016–2019, coordinated by the University of Innsbruck) that built a Virtual Research Environment for the automated recognition, transcription, indexing and enrichment of handwritten archival documents. Its main outcome was the Transkribus platform for Handwritten Text Recognition.

Visit project ↗

tranScriptorium — an EU FP7 project (2013–2015) developing cost-effective solutions for the indexing, search and full transcription of historical handwritten document images using holistic, segmentation-free Handwritten Text Recognition (HTR): image preprocessing, keyword indexing/search, and interactive-predictive tools for computer-assisted transcription. Its research fed into the Transkribus platform and the successor Horizon 2020 project READ.

Visit project ↗

THALES-3DOR (MIS 379516) — a Greek national research project on the retrieval of static and spatiotemporal 3D objects: global/rigid shape descriptors, 3D object alignment, retrieval from 2D queries, and 3D video / mesh-sequence retrieval. Led by Ioannis Pratikakis (Democritus University of Thrace) with T. Theoharis (University of Athens), funded under the THALES (ΘΑΛΗΣ) action of the NSRF “Education and Lifelong Learning” programme, co-financed by the European Social Fund and Greek national funds.

Visit project ↗

Provision of research services for IMRA Europe S.A.S. (2014–2015) — a research-services collaboration evaluating the discriminative power of shape descriptors for pedestrian / non-pedestrian recognition from 3D LIDAR data, studying the influence of point-cloud density and delivering a proof-of-concept. The work contributed to the 2018 Computer Vision and Image Understanding paper on statistical shape analysis for pedestrian recognition from LIDAR.

Visit project ↗

A web-based e-training platform for Extended Human Motion Investigation in Orthopedics — a Lifelong Learning Programme / Leonardo da Vinci “Transfer of Innovation” project (2012–2013) that created a web-based “Virtual Training & Communication Center” for human-motion analysis in orthopedics, with E-Learning, E-Communicating and E-Mentoring components bridging engineering and medicine.

Visit project ↗