Please use this identifier to cite or link to this item:
https://hdl.handle.net/10316/93988
DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Fernandes, Gabriel Falcão Paiva | - |
dc.contributor.author | Barreiros, João Carlos da Costa | - |
dc.date.accessioned | 2021-03-29T22:16:41Z | - |
dc.date.available | 2021-03-29T22:16:41Z | - |
dc.date.issued | 2020-12-17 | - |
dc.date.submitted | 2021-03-29 | - |
dc.identifier.uri | https://hdl.handle.net/10316/93988 | - |
dc.description | Dissertação de Mestrado Integrado em Engenharia Electrotécnica e de Computadores apresentada à Faculdade de Ciências e Tecnologia | - |
dc.description.abstract | Feature extraction of high-resolution images is a challenging procedure in low-power signal processing applications. This thesis describes how to optimize and efficiently parallelize the scale-invariant feature transform (SIFT) feature detection algorithm and maximize the use of bandwidth on the GPUsubsystem. Together with the minimization of data communications between host and device, the successful parallelization of all the main kernels used in SIFT allowed a global speedup in high-resolution images above 78x while being more than an order of magnitude energy efficient (FPS/W) than its serial counterpart. From the 3 GPUs tested, the low-power GPU has shown superior energy efficiency -- 44 FPS/W. | por |
dc.description.abstract | Feature extraction of high-resolution images is a challenging procedure in low-power signal processing applications. This thesis describes how to optimize and efficiently parallelize the scale-invariant feature transform (SIFT) feature detection algorithm and maximize the use of bandwidth on the GPUsubsystem. Together with the minimization of data communications between host and device, the successful parallelization of all the main kernels used in SIFT allowed a global speedup in high-resolution images above 78x while being more than an order of magnitude energy efficient (FPS/W) than its serial counterpart. From the 3 GPUs tested, the low-power GPU has shown superior energy efficiency -- 44 FPS/W. | eng |
dc.language.iso | eng | - |
dc.rights | openAccess | - |
dc.rights.uri | http://creativecommons.org/licenses/by-nc/4.0/ | - |
dc.subject | Feature extraction | por |
dc.subject | Scale-invariant feature transform | por |
dc.subject | GPGPU | por |
dc.subject | CUDA | por |
dc.subject | Parallel Programming | por |
dc.subject | Feature extraction | eng |
dc.subject | Scale-invariant feature transform | eng |
dc.subject | GPGPU | eng |
dc.subject | CUDA | eng |
dc.subject | Parallel Programming | eng |
dc.title | Fast Scale-Invariant Feature Transform on GPU | eng |
dc.title.alternative | Fast Scale-Invariant Feature Transform on GPU | por |
dc.type | masterThesis | - |
degois.publication.location | DEEC | - |
degois.publication.title | Fast Scale-Invariant Feature Transform on GPU | eng |
dc.peerreviewed | yes | - |
dc.identifier.tid | 202686574 | - |
thesis.degree.discipline | Engenharia Electrotécnica e de Computadores | - |
thesis.degree.grantor | Universidade de Coimbra | - |
thesis.degree.level | 1 | - |
thesis.degree.name | Mestrado Integrado em Engenharia Electrotécnica e de Computadores | - |
uc.degree.grantorUnit | Faculdade de Ciências e Tecnologia - Departamento de Eng. Electrotécnica e de Computadores | - |
uc.degree.grantorID | 0500 | - |
uc.contributor.author | Barreiros, João Carlos da Costa::0000-0003-4667-4847 | - |
uc.degree.classification | 18 | - |
uc.degree.presidentejuri | Barreto, João Pedro de Almeida | - |
uc.degree.elementojuri | Fernandes, Gabriel Falcão Paiva | - |
uc.degree.elementojuri | Lobo, Jorge Nuno de Almeida e Sousa Almada | - |
uc.contributor.advisor | Fernandes, Gabriel Falcão Paiva::0000-0001-9805-6747 | - |
item.fulltext | Com Texto completo | - |
item.openairecristype | http://purl.org/coar/resource_type/c_18cf | - |
item.grantfulltext | open | - |
item.languageiso639-1 | en | - |
item.openairetype | masterThesis | - |
item.cerifentitytype | Publications | - |
crisitem.advisor.researchunit | IT - Institute of Telecommunications | - |
crisitem.advisor.orcid | 0000-0001-9805-6747 | - |
Appears in Collections: | UC - Dissertações de Mestrado |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
DissertacaoJoaoBarreiros_23 versão definitiva.pdf | 12.74 MB | Adobe PDF | View/Open |
Google ScholarTM
Check
This item is licensed under a Creative Commons License