Multi sensors data fusion and Object detection algorithms for in-disaster scene situation awareness

Autor: Sorrentino, Viola, Di Massa, Vincenzo, Guerri, Marco, Mando, Gianluca, Mueller, Nikolas, Symeonidis, Ioannis, Karlsson, Patrik, Maliou, Iliana
Jazyk: angličtina
Rok vydání: 2021
Předmět:
DOI: 10.5281/zenodo.5833882
Popis: This Deliverable called “Multi sensors data fusion and Object detection algorithms for in-disaster scene situation awareness” is the result of the Task 3.4 of Search&Rescue Project where the design and development of Obstacle Detection System (ODS) is described. In particular the following topic are introduced: the scope of a Robotic solution with an ODS onboard, the importance of the ODS within the S&R Project and the specific advantages to have a multi sensor data fusion algorithm. Then the System Overview is presented treating the related references, the Machine learning, AI, the Processing Unit, the Implementation notes and finally their integration in the selected S&R rescue Robot. In the following chapters the main topics are described: the architecture of Obstacle Detection System with smart sensors, special synchronisation and depth estimation; the implementation of Obstacle Detection Algorithms and finally the Fusion and Tracking operations. At the end of the deliverable there are references to the Robot System Integration (with recall to the D5.4) and the related Verification and Validation of the System.
Databáze: OpenAIRE