Master's Dissertation
DOI
https://doi.org/10.11606/D.45.2019.tde-29062019-223239
Document
Author
Full name
Jefferson Serafim Ascaneo
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2017
Supervisor
Committee
Hirata Junior, Roberto (President)
Kogler Junior, Joao Eduardo
Lopes, Rubens Mendes
Title in Portuguese
Processamento de imagens de holografia digital para o estudo de organismos planctônicos
Keywords in Portuguese
Holografia
Plancton
Processamento de imagens
Processamento digital de sinais
Visão computacional
Abstract in Portuguese
O estudo de plancton e importante para estimar a biodiversidade em ambientes marinhos, auxiliando na compreensao de sua dinamica e do impacto da atividade humana. Neste trabalho, estudamos as etapas necessarias para a observacao de organismos planctonicos utilizando um sistema holografico digital em linha. A holografia digital permite que as imagens sejam numericamente focalizadas apos sua aquisicao atraves de algoritmos de reconstrucao holografica, o que remove a limitacao da baixa profundidade de foco presente em metodos mais tradicionais, e permite a observacao de um volume maior de agua. E feita uma analise das limitacoes teoricas deste sistema holografico em termos das frequencias presentes no holograma e na sua reconstrucao, observando-se as consequencias para a escolha e implementacao dos algoritmos de reconstrucao. Tambem sao considerados alguns detalhes de implementacao que decorrem da natureza finita e discreta dos algoritmos. No processamento dos hologramas, os organismos sao automaticamente localizados segmentando-se regioes de interesse por meio de uma reconstrucao volumetrica. Estas regioes sao reconstruidas a diversas distancias e e feita a escolha de um plano focal utilizando uma medida de foco. Diversas medidas sao comparadas em imagens obtidas experimentalmente para determinar a mais adequada na focalizacao de organismos planctonicos. Por fim, descrevemos a aplicacao web desenvolvida para o processamento distribuido de imagens holograficas digitais de plancton.
Title in English
Processing of digital holography images for the study of planktonic organisms
Keywords in English
Computer vision
Digital signal processing
Holography
Image processing
Plankton
Abstract in English
The study of plankton is important for estimating biodiversity in marine environments, helping in the understanding of its dynamics and the impact of human activity. In this work, we study the necessary steps for the observation of planktonic organisms using an inline digital holographic system. Digital holography allows images to be numerically focused after its acquisition through holographic reconstruction algorithms, which removes the limitation of low depth of focus present in more traditional methods, and allows the observation of a bigger volume of water. We present an analysis of the theoretical limitations of this holographic system concerning the frequencies present in the hologram and its reconstruction, noting the consequences for the choice and implementation of reconstruction algorithms. Also considered are some implementation details that stem from the finite and discrete nature of the algorithms. In the processing of holograms, organisms are automatically located by segmenting regions of interest in a volumetric reconstruction. These regions are reconstructed at numerous distances and the choice of a focal plane is made using the values of a focus measure. Various measures are compared in experimentally obtained images to determine the most adequate for focusing planktonic organisms. Finally, we describe a web application developed for distributed processing of digital holographic images of plankton.
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
2019-07-10