Significant edges in the case of non-stationary Gaussian noise

DOI10.1016/J.PATCOG.2007.02.015MaRDI QIDQ2643916zbMATH OpenOpenAlexFDO

Authors Romain Abraham, Agnès Desolneux, S. Li-Thiao-Te, Isabelle Abraham

Publication date 27 August 2007

Published in Pattern Recognition (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/math/0606219

zbMATH Keywords

edge detectioninverse problemsignificant edgesstatistical hypothesis testing

Mathematics Subject Classification ID

Pattern recognition, speech recognition (68T10)

Abstract: In this paper, we propose an edge detection technique based on some local smoothing of the image followed by a statistical hypothesis testing on the gradient. An edge point being defined as a zero-crossing of the Laplacian, it is said to be a significant edge point if the gradient at this point is larger than a threshold $s(eps)$ defined by: if the image $I$ is pure noise, then $$. In other words, a significant edge is an edge which has a very low probability to be there because of noise. We will show that the threshold $s(eps)$ can be explicitly computed in the case of a stationary Gaussian noise. In images we are interested in, which are obtained by tomographic reconstruction from a radiograph, this method fails since the Gaussian noise is not stationary anymore. But in this case again, we will be able to give the law of the gradient conditionally on the zero-crossing of the Laplacian, and thus compute the threshold $s(eps)$. We will end this paper with some experiments and compare the results with the ones obtained with some other methods of edge detection.

Recommendations

• Detection of edges from nonuniform Fourier data
• Statistical behavior of edge detectors
• scientific article; zbMATH DE number 3890451
• Generalization of Canny-Deriche filter for detection of noisy exponential edge
• Detecting edges from non-uniform Fourier data using Fourier frames

Cites work

• A Multiscale Algorithm for Image Segmentation by Variational Method
• Edge detection by Helmholtz principle
• Extracting meaningful curves from images
• scientific article; zbMATH DE number 3115573 (Why is no real title available?)
• scientific article; zbMATH DE number 3984248 (Why is no real title available?)
• scientific article; zbMATH DE number 3658788 (Why is no real title available?)
• scientific article; zbMATH DE number 408258 (Why is no real title available?)

Cited in

(5)

• When the a contrario approach becomes generative
• Statistical behavior of edge detectors
• scientific article; zbMATH DE number 1121142 (Why is no real title available?)
• A new Poisson noise filter based on weights optimization
• A gravitational approach to edge detection based on triangular norms

This page was built for publication: Significant edges in the case of non-stationary Gaussian noise

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2643916)