| KEYWORD |
Study and development of imagle elaboration software oriented to automotive applications
keywords GPU, PEDESTRIAN RECOGNITION, DEPENDABILITY, SAFETY
Reference persons MATTEO SONZA REORDA, LUCA STERPONE
Research Groups ELECTRONIC CAD & RELIABILITY GROUP - CAD
Thesis type RESEARCH THESIS
Description The goal of the thesis is the realization of an application for the image elaboration
recorded by a cam-corder vehicle oriented to the pedestrian recognition. This application will be
devloped using General Purpose Graphic Processign Units (GPGPUs) manufactured by NVIDIA.
In particular using the state-of-the-art GPU devices for embedded systems (NVIDIA JETSON TK1).
The application must be developed considering fault mitigation methodology and automotive
dependability constraints (using automotive standards) and it will be deployed in collaboration
with NVIDIA and the Porto Alegre University (UFRGS).
During the thesis it is expected an experimental activity within the world's most used radiation
particle accelerators: Los Alamos (USA) and ISIS (UK). They will be used to emulate fault conditions
thanks to the radiation particle generation.
Required skills C/C++ programming language
Deadline 18/04/2015
PROPONI LA TUA CANDIDATURA