Distance transform network for shape analysis

Ribas, Lucas Correia; Neiva, Mariane Barros; Bruno, Odemir Martinez

doi:10.1016/j.ins.2018.08.038

Cited by 21 publications

(9 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Redes de Transformada da Distância para Análise de Formas (DTN). Abordagem para a extrac ¸ão de características de formas que foi publicada em [Ribas et al 2019]. Diferentemente do método proposto por [Backes et al 2009], neste trabalho, a transformada da distância Euclidiana (em inglês, Euclidean Distance Transform (EDT)) é aplicada na imagem de contorno e, em seguida, os pixels que pertencem a um dado raio de dilatac ¸ão r são modelados como uma RC.…”

Section: Análise De Formasunclassified

Aprendizado de representações e caracterização de redes complexas com aplicações em visão computacional

Ribas

Bruno

2022

Anais Do XXXV Concurso De Teses E Dissertações (CTD 2022)

Self Cite

View full text Add to dashboard Cite

Este trabalho apresenta um resumo da tese de doutorado que tem como objetivo investigar e propor novas técnicas de modelagem e caracterização de redes complexas focando na aplicação em problemas de visão computacional. Em relação à modelagem, foi estudada uma abordagem eficaz e otimizada para mapear as imagens e vídeos de texturas em redes complexas direcionadas. A respeito da caracterização, novas formas de caracterizar as redes complexas foram investigadas, com destaque para o uso de redes neurais randomizadas para aprendizado de representações, resultando em vários métodos propostos para análise de texturas em níveis de cinza, coloridas e dinâmicas e análise de formas. Resultados promissores foram alcançados pelos métodos desenvolvidos em comparação aos métodos da literatura em tarefas de classificação de imagens/vídeos usando vários conjuntos de dados que são referência. Adicionalmente, para avaliar o potencial dos métodos desenvolvidos, foram investigadas cinco aplicações em problemas reais nas áreas de biologia, botânica, físico-química e medicina, alcançando resultados interessantes e contribuindo para o desenvolvimento destas áreas.

show abstract

Section: Análise De Formasunclassified

Aprendizado de representações e caracterização de redes complexas com aplicações em visão computacional

Ribas

Bruno

2022

Anais Do XXXV Concurso De Teses E Dissertações (CTD 2022)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Leaf plant classification is used in their experiment to validate the effectiveness of their complex network based shape descriptors. Inspired by the work in [46], Ribas et al [47] proposed a distance transform network, which incorporated the network with a Euclidean distance transform, for shape analysis and applied their method for leaf recognition.…”

Section: Related Workmentioning

confidence: 99%

Chord Bunch Walks for Recognizing Naturally Self-Overlapped and Compound Leaves

Wang

Gao

Sun

et al. 2019

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Effectively describing and recognizing leaf shapes under arbitrary variations, particularly from a large database, remains an unsolved problem. In this research, we attempted a new strategy of describing leaf shapes by walking and measuring along a bunch of chords that pass through the shape. A novel chord bunch walks (CBW) descriptor is developed through the chord walking behaviour that effectively integrates the shape image function over the walked chord to reflect both the contour features and the inner properties of the shape. For each contour point, the chord bunch groups multiple pairs of chords to build a hierarchical framework for a coarse-to-fine description that can effectively characterize not only the subtle differences among leaf margin patterns but also the interior part of the shape contour formed inside a self-overlapped or compound leaf. Instead of using optimal correspondence based matching, a Log-Min distance that encourages one-to-one correspondences is proposed for efficient and effective CBW matching. The proposed CBW shape analysis method is invariant to rotation, scaling, translation, and mirror transforms. Five experiments, including image retrieval of compound leaves, image retrieval of naturally self-overlapped leaves, and retrieval of mixed leaves on three large scale datasets, are conducted. The proposed method achieved large accuracy increases with low computational costs over the state-of-the-art benchmarks, which indicates the research potential along this direction.

show abstract

“…Currently, methods based on complex networks theory have attracted significant attention of the computer vision community, due to the promising results and their capacity to represent the relationships among structural properties of texture [15,38,1]. The complex network techniques represent the texture image as a network and extract measures capable of characterizing precisely the structural patterns of texture modeled.…”

Section: Introductionmentioning

confidence: 99%

A Fractal-Based Approach to Network Characterization Applied to Texture Analysis

Ribas

Manzanera

Bruno

2019

Computer Analysis of Images and Patterns

View full text Add to dashboard Cite

This work proposes a new method for texture analysis that combines fractal descriptors and complex network modeling. At first, the texture image is modeled as a network. Then, the network is converted into a surface where the Cartesian coordinates and the vertex degree is mapped into a 3D point in the surface. Then, we calculate a description vector of this surface using a method inspired by the Bouligand-Minkowski technique for estimating the fractal dimension of a surface. Specifically, the descriptor corresponds to the evolution of the volume occupied by the dilated surface, when the radius of the spherical structuring element increases. The feature vector is given by the concatenation of the volumes of the dilated surface for different radius values. Our proposal is an enhancement of the classic complex networks descriptors, where only the statistical information was considered. Our method was validated on four texture datasets and the results reveal that our method leads to highly discriminative textural features.

show abstract

Distance transform network for shape analysis

Cited by 21 publications

References 32 publications

Aprendizado de representações e caracterização de redes complexas com aplicações em visão computacional

Aprendizado de representações e caracterização de redes complexas com aplicações em visão computacional

Chord Bunch Walks for Recognizing Naturally Self-Overlapped and Compound Leaves

A Fractal-Based Approach to Network Characterization Applied to Texture Analysis

Contact Info

Product

Resources

About