• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.3.2020.tde-05042021-150339
Document
Author
Full name
Carlos Fabbri Junior
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2020
Supervisor
Committee
Santos, Josemir Coelho (President)
Penedo, Sergio Ricardo Master
Sakamoto, João Marcos Salvi
Title in Portuguese
Recuperação de imagens multiespectrais por sensoriamento compressivo com o uso de câmeras de pixel único e algoritmos convexos de programação linear
Keywords in Portuguese
Amostragem
Imagens (Recuperação)
Processamento de imagens
Sensoriamento compressivo
Abstract in Portuguese
O presente trabalho mostra uma nova metodologia para a aquisição de imagens do espectro visível, IR e UV usando um número significativamente menor de amostras do que a teoria convencional de Shannon-Nyquist recomenda. Essa nova metodologia é baseada em uma teoria inovadora e revolucionária chamada de Compressed Sensing ou sensoriamento compressivo. Ela propõe um novo método de captura das informações essenciais da imagem ou objeto sendo amostrados, baseado no conhecimento de que essas informações são esparsas em uma determinada base de representação da informação. Duas características fundamentais para esse feito são a esparsidade da imagem e a incoerência entre a base de representação e a base de medida do objeto. Algoritmos de programação linear foram desenvolvidos para reconstruir a imagem original a partir das amostras obtidas, com alto grau de sucesso. O presente trabalho explica o funcionamento de um destes algoritmos etapa por etapa. Exemplos de imagens reconstruídas usando um dos algoritmos propostos fazem parte do presente trabalho.
Title in English
Compressive sensing multispectral images recovering by using single pixel camera and linear programming convex algorithms.
Keywords in English
Compressed sensing
Compressive sampling Sparsity
Dimensionality reduction
Image processing
Image recovery
Sampling
Abstract in English
The present work shows a new metodology for the capture of IR, UV and visible images using a significantly smaller number of samples as dictated by the main Shannon-Nyquist sampling theory. This new metodology is based on a revolutionary and innovative theory called compressed sensing or compressive sampling. This theory proposes a new method for capturing the main information of the image or the object being sampled, based on the knowledge that these informations are sparse on a specific information coding base. Two essential caractheristics for this feature are sparsity of the image and the incoherence between the coding base and the measurement base. Linear programming algorithms were developed to reconstruct the original image from the samples obtained, with high level of success. The present work explains step by step the working principle of such algorithm. Examples of reconstructed images using the proposed algorithms are included in this work.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2021-04-06
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.