Ridges and Valleys Detection in Images Using Difference of Rotating Half Smoothing Filters

Abstract

In this paper we propose a new ridge/valley detection method in images based on the difference of rotating Gaussian semi filters. The novelty of this approach resides in the mixing of ideas coming both from directional filters and DoG method. We obtain a new ridge/valley anisotropic DoG detector enabling very precise detection of ridge/valley points. Moreover, this detector performs correctly at crest lines even if highly bended, and is precise on junctions. This detector has been tested successfully on various image types presenting difficult problems for classical ridges/valleys detection methods.

Perceptual Color Image Smoothing via a New Region-Based PDE Scheme

“…This technique is very attractive because of the simplicity of the filters and the intelligibility of their functioning, which helps finding perceptuel features close to human vision in images. These rotating filters have been used in a number of applications such as edge and ridges detection, texture suppression or image regularization [17,14,13,12,15]. In the method presented in this paper, a smoothing rotating filter first enables classification between edges and homogeneous regions, so that either isotropic or anisotropic smoothing can be used.…”

“…2) in order to build a signal s, which is a function of a rotation angle θ and the underlying signal. As shown in [17,14,12,13], smoothing with rotating filters implies that the image is smoothed with a bank of rotated anisotropic Gaussian half-kernels:…”

“…This paper introduces a rotating filter (inspired by [17,14,13]) able to detect homogenous regions and edges regions, even in images containing heavy noise. Then, it presents an anisotropic edge detector capable of defining two directions for pixels belonging to edges.…”

A New Region-Based Pde for Perceptual Image Restoration

“…In this paper, we present a rotating filter (inspired by (Montesinos and Magnier, 2010), (Magnier et al, 2011c) and (Magnier et al, 2011b)) able to detect homogenous regions and edges regions, even in highly noisy images. Then, we present an anisotropic edge detector which defines two directions for pixels belonging to edges.…”

“…1) in order to build a signal s which is a function of a rotation angle θ and the underlying signal. As shown in (Montesinos and Magnier, 2010), (Magnier et al, 2011c) and (Magnier et al, 2011b), smoothing with rotating filters means that the image is smoothed with a bank of rotated anisotropic Gaussian half kernels:…”