A Decomposition and Noise Removal Method Combining Diffusion Equation and Wave Atoms for Textured Images

Casaca, Wallace; Boaventura, Maurílio

doi:10.1155/2010/764639

Cited by 11 publications

(4 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The corresponding fractional‐order variational models have become our targets. We can also consider extending the proposed fractional‐order PDE to other PDEs [36–38] for texture image denoising.…”

Section: Discussionmentioning

confidence: 99%

Image decomposition and denoising using fractional‐order partial differential equations

Bai

Feng

2020

IET image process

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Image decomposition and denoising using fractional‐order partial differential equations

Bai

Feng

2020

IET image process

View full text Add to dashboard Cite

“…This proposed method derives from the so-called spectral clustering methods (reported in [Casaca 2014, Casaca et al 2013b) and it combines cartoon-texture decomposition [Casaca and Boaventura 2010], similarity metrics [Casaca et al 2013b], and spectral graph theory [Chung 1997] into a unified framework (see Figure 1). The proposed approach holds attractive properties such as awareness to noise and texture, accuracy in detecting image edges, low computational cost and a reduced number of human intervention (see Figure 2 for a few illustrations).…”

Section: Spectral Image Segmentationmentioning

confidence: 99%

Graph Laplacian for Spectral Clustering and Seeded Image Segmentation

Casaca¹,

Taubin²,

Nonato³

2015

Anais Do XXVIII Concurso De Teses E Dissertações (CTD 2015)

View full text Add to dashboard Cite

Interactive segmentation methods have gained much attention lately, specially due to their good performance in segmenting complex images and easy utilization. However, most interactive segmentation algorithms rely on sophisticated mathematical formulations whose effectiveness highly depends on the kind of image to be processed. In fact, sharp adherence to the contours of image segments, uniqueness of solution, high computational burden, and extensive user interaction are some of the weaknesses of most existing methods. In this thesis we proposed two novel interactive image segmentation techniques that sort out the issues raised above. The proposed methods rely on Laplace operators, spectral graph theory, and optimization schemes towards enabling highly accurate segmentation tools which demand a reduced amount of user interaction while still being mathematically simple and computationally efficient. The good performance of our segmentation algorithms is attested by a comprehensive set of comparisons against representative state-of-the-art methods. As additional contribution, we have also proposed two new algorithms for inpainting and photo colorization, both of which rely on the accuracy of our segmentation apparatus

show abstract

“…), representada no sistema de cores RGB, sendo ela submetida a um filtro de suavização gaussiano (com σ = 1) a fim de gerar uma imagem suavizada I suave(Figura 1(b)). Na sequência, a imagem componente I po(Figura 1(c)) contendo os padrões oscilatórios da imagem original Ié obtida pela equação I po = I − I suave , tal como realizado pela decomposição de imagens apresentada em[3].…”

unclassified

Um Método Não-Supervisionado de Detecção de Oclusões Textuais para Imagens de Sensoriamento Remoto

Basso¹,

Colnago²,

Casaca³

2020

Proceeding Series of the Brazilian Society of Computational and Applied Mathematics

View full text Add to dashboard Cite

(PDI) têm sido extensivamente empregadas em imagens de sensoriamento remoto. Nesse contexto, classificar alvos ou extrair características de interesse podem se tornar tarefasárduas devidoà grande quantidade de oclusões que podem surgir nas imagens como, por exemplo, sombras de nuvens, textos sobrepostos, falhas na captura de dados por sensores, etc. Em vista disso, este trabalho apresenta uma nova metodologia para detecção automática de oclusões do tipo textual em imagens de sensoriamento remoto, a qual combina técnicas de segmentação morfológica, decomposição de imagens e estratégias de limiarização. A solução proposta considera filtros gaussianos para suavizar a imagem de forma a obter uma decomposição da mesma nas bandas "cartoon" e "texture" (padrões oscilatórios). Em seguida, essa bandas são processadas e binarizadas a partir do método de Otsu, gerando assim uma imagem auxilar a qualé submetida a operações morfológicas, de modo a guiar a detecção textual na imagem original. Testes experimentais em imagens reais remotamente sensoriadas demonstram que o método proposto atinge elevada precisão na tarefa de detecção desse tipo de alvo, devolvendo imagens comáreas textuais segmentadas que podem ser utilizadas para aumentar a acurácia de outras aplicações clássicas daárea como classificação e extração de feições de interesse. Palavras-chave. Análise e processamento de imagens, Morfologia matemática, Detecção automática, Sensoriamento remoto, Filtros de suavização.

show abstract

A Decomposition and Noise Removal Method Combining Diffusion Equation and Wave Atoms for Textured Images

Cited by 11 publications

References 31 publications

Image decomposition and denoising using fractional‐order partial differential equations

Image decomposition and denoising using fractional‐order partial differential equations

Graph Laplacian for Spectral Clustering and Seeded Image Segmentation

Um Método Não-Supervisionado de Detecção de Oclusões Textuais para Imagens de Sensoriamento Remoto

Contact Info

Product

Resources

About