Ημερομηνία: Παρασκευή 6 Ιουλίου, 2012 Ώρα: 12:00 - 14:00
Τοποθεσία: Aίθουσα Συναντήσεων "Αλκιβιάδης Χ. Παγιατάκης", ΙΤΕ, Ηράκλειο, Κρήτη
Host: Αργυρός Αντώνιος
Motion planning for mobile robots constitutes a domain of research where several disciplines meet, ranging from artificial intelligence and machine learning to robot perception and computer vision. In view of the plurality of related applications such as planetary exploration, military, search and rescue, agriculture, mining and off-road exploration, the aim of this talk will be to present the current state-of-the-art in the domain of 3D terrain perception and analysis that is employed at a preceding stage as a means to effectively and efficiently guide the task of UGV motion planning. We identify that in the epicenter of all related methodologies, 3D terrain information is used that is acquired from LIDAR, stereo range data, colour or other sensory data and occasionally combined with static or dynamic vehicle models that express the interaction of the robotic vehicle with the terrain. We form a taxonomy of the explored directions that is comprehensive of the dispersed contributions in the field as well as inspiring in addressing the challenges and envisioning the future directions.
Panagiotis Papadakis received the University degree in Informatics and Telecommunications in 2005 and the D. Phil. degree in Information Technology from the National University of Athens, Greece, in 2009. Since 2010, he is a post-doctoral researcher at the ALCOR - Vision, Perception and Cognitive Robotics Laboratory - of the University of Rome "La Sapienza", Italy, where he is pursuing research mainly in the domain of **perception and planning for **Search & Rescue Robotics. His research interests span the fields of computer vision, machine learning and robotics with a special focus on robot perception, content-based 3D object retrieval, pose normalization, 2D/3D segmentation and action recognition. He is also the primary inventor of two international patents on content-based 3D object retrieval.