Random Germs and Stochastic Watershed for Unsupervised Multispectral Image Segmentation

Noyel, Guillaume; Angulo, Jesús; Jeulin, Dominique

doi:10.1007/978-3-540-74829-8_3

Cited by 16 publications

(33 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The stochastic watershed was proved to be efficient for unsupervised segmentation (Noyel et al, 2007;Faessel and Jeulin, 2010). For multiscale images, the full granulometry of the image is used Gillibert and Jeulin (2011b).…”

Section: Multiscale Image Segmentationmentioning

confidence: 99%

3d Reconstruction and Analysis of the Fragmented Grains in a Composite Material

Gillibert¹,

Jeulin²

2013

Image Anal Stereol

Self Cite

View full text Add to dashboard Cite

X-ray microtomography from solid propellant allows studying the microstructure of fragmented grains in damaged samples. A new reconstruction algorithm of fragmented grains for 3D images is introduced. Based on a watershed transform of a morphological closing of the input image, the algorithm can be used with different sets of markers. Two of them are compared. After the grain reconstruction, a multiscale segmentation algorithm is used to extract each fragment of the damaged grains. This allows an original quantitative study of the fragmentation of each grain in 3D. Experimental results on X-ray microtomographic images of a solid propellant fragmented under compression are presented and validated.

show abstract

Section: Multiscale Image Segmentationmentioning

confidence: 99%

3d Reconstruction and Analysis of the Fragmented Grains in a Composite Material

Gillibert¹,

Jeulin²

2013

Image Anal Stereol

Self Cite

View full text Add to dashboard Cite

show abstract

“…The potential function V x c ( ) means the potential between the two pixel labels of the clique in the neighbour− hood. If the two labels is represented by q i j ( , ) and q i j ( , ) 1 …”

Section: Gaussian Mixture Model and Mrfmentioning

confidence: 99%

“…The remote sensing image contains plenty of information. But it is difficult to cluster the remote sensing images because of their fuzzy details [1,2].…”

Section: Introductionmentioning

confidence: 99%

Mixture model and Markov random field-based remote sensing image unsupervised clustering method

Hou

Yang²,

Rao

et al. 2011

Opto-Electronics Review

View full text Add to dashboard Cite

In this paper, a novel method for remote sensing image clustering based on mixture model and Markov random field (MRF) is proposed. A remote sensing image can be considered as Gaussian mixture model. The image clustering result corresponding to the image label field is a MRF. So, the image clustering procedure is transformed to a maximum a posterior (MAP) problem by Bayesian theorem. The intensity difference and the spatial distance between the two pixels in the same clique are introduced into the traditional MRF potential function. The iterative conditional model (ICM) is employed to find the solution of MAP. We use the max entropy criterion to choose the optimal clustering number. In the experiments, the method is compared with the traditional MRF clustering method using ICM and simulated annealing (SA). The results show that this method is better than the traditional MRF model both in noise filtering and miss-classification ratio.

show abstract

“…The obtained contours are more regular and significant than these associated to the deterministic watershed. Probabilistic watershed was then extended to hyperspectral images by Noyel, Angulo and Jeulin [5].…”

Section: Introductionmentioning

confidence: 99%

“…For hyperspectral images, a pdf is built for each channel of the image and the flooding function is the weighted sum of the pdf of the channels. This function, called a marginal probability density function, is based on spatial information [5].…”

Section: Introductionmentioning

confidence: 99%