• 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
 
 
Doctoral Thesis
DOI
https://doi.org/10.11606/T.45.2000.tde-20220712-115205
Document
Author
Full name
Nina Sumiko Tomita Hirata
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2000
Supervisor
Title in Portuguese
Projeto automático de operadores: explorando conhecimentos a priori
Keywords in Portuguese
Computação Gráfica
Processamento De Imagens
Abstract in Portuguese
A morfologia matemática vem sendo largamente utilizada para processamento e análise de imagens digitais. O projeto de operadores morfológicos é em geral realizado de forma heurística. Devido à dificuldade inerente a este procedimento, técnicas de projeto automático são de grande importância e interesse. Várias abordagens neste sentido vêm sendo propostas, dentre elas técnicas que projetam operadores a partir de exemplos de treinamento (obtidos de amostras de imagens observadas-ideais) que representam de forma simples a transformação desejada pelo usuário. Tomando uma técnica de projeto de operadores baseada no modelo de aprendizado PAC (do inglês, 'Probably Approximately Correct') como ponto de partida, investigamos de forma geral algumas das limitações dessas abordagens. Com base nessa investigação, estudamos o projeto de W-operadores, colocando ênfase sobre questões relacionadas com a precisão de operadores projetados a partir de uma quantidade limitada de exemplos de treinamento. Os frutos deste estudo, apresentamos neste trabalho, são técnicas que exploram conhecimentos sobre o problema que desejamos resolver para projetar operadores mais precisos e algoritmos eficientes para implementar as mesmas. Soluções para problemas reais de processamento de imagens ilustram a aplicação das técnicas propostas
Title in English
not available
Abstract in English
Mathematical morphology is being widely used in image processing and analysis. Designing morphological operators is usually done by heuristic methods. However, due to the inherent difficulty of such procedures, automatic design techniques are of increasing interest. In recent years, several approaches for the automatic design of morphological operators have been proposed. Some of them are based on learning from training examples (sampled from observed-ideal pairs of images representing the desired image processing mapping). Starting from a technique based on PAC (probably Approximately Correct) learning model, we investigate some limitations of those approaches. From this investigation, we study the design ofW-operators emphatizing questions related with precision of operators designed from a limited number of training examples. The results of this study, presented in this work, are techniques which exploit knowledge (about the image processing problem being solved) in order to design more accurate operators and efficient algorithms for implementing them. Solutions for some real image processing problems are given to illustrate the application of the proposed techniques
 
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
2022-07-13
 
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.