Active Unsupervised Texture Segmentation on a Diffusion Based Feature Space

نویسندگان

  • Mikaël Rousson
  • Thomas Brox
  • Rachid Deriche
چکیده

In this report, we propose a novel and e cient approach for active unsurpervised texture segmentation. First, we show how we can extract a small set of good features for texture segmentation based on the structure tensor and nonlinear di usion. Then, we propose a variational framework that allows to incorporate these features in a level set based unsupervised segmentation process that adaptively takes into account their estimated statistical information inside and outside the region to segment. Unlike features obtained by Gabor lters, our approach naturally leads to a signi cantly reduced number of feature channels. Thus, the supervised part of a texture segmentation algorithm, where the choice of good feature channels has to be learned in advance, can be omitted, and we get an e cient solution for unsupervised texture segmentation. The actual segmentation process based on the new features is an active and adaptative contour model that estimates dynamically probability density functions inside and outside a region and produces very convincing results. It is implemented using a fast level set based active contour technique and has been tested on various real textured images. The performance of the approach is favorably compared to recent studies. Key-words: Level Set Theory, Texture Segmentation, Adaptative Image Segmentation, Nonlinear Di usion, Structure Tensor. Segmentation active et non supervisée d'images texturées à l'aide d'une di usion non linéaire du tenseur de structure Résumé : Dans ce rapport, nous proposons une nouvelle approche pour la segmentation active d'images texturées. Tout d'abord, nous présentons une méthode d'extraction d'un nombre restreint de composantes pour caractériser l'information de texture présente dans l'image à segmenter. Ce processus est basé sur le tenseur de structure et la di usion non-linéaire. Ensuite, nous proposons un cadre variationnel a n d'incorporer ces di érentes caractérisitiques dans un processus de segmentation adaptatif et non supervisé, basé sur les ensembles de niveaux. Contrairement aux approches utilisant des ltres de Gabor pour extraire l'information texture, notre approche fournit naturellement un nombre réduit de composantes. Ainsi, la partie supervisée pour la segmentation d'images texturées, où le choix des bonnes caractéristiques est issu d'un processus d'apprentissage, peut être évitée et nous obtenons une solution e cace pour une segmentation non-supervisée d'images texturées. Basé sur ces nouvelles composantes, nous proposons un processus de segmentation actif et adaptatif où les densités de probabilités à l'intérieur et a` l'extérieur du contour sont estimées de manière dynamique. Nous utilisons une technique rapide basée sur les ensembles de niveaux pour la mise en oeuvre. Pour nir, mous présentons des résultats de validation sur diverses images réelles et nous les comparons avec succès à ceux obtenus à partir d'études récentes. Mots-clés : Théorie des courbes de niveaux, segmentation d'images texturées, segmentation adaptative d'images, di usion non-linéaire, tenseur de structure. Active Unsupervised Texture Segmentation on a Di usion Based Feature Space 3

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Unsupervised Texture Segmentation via Applying Geodesic Active Regions to Gaborian Feature Space

In this paper, we propose a novel variational method for unsupervised texture segmentation. We use a Gabor filter bank to extract texture features. Some of the filtered channels form a multidimensional Gaborian feature space. To avoid deforming contours directly in a vector-valued space we use a Gaussian mixture model to describe the statistical distribution of this space and get the boundary a...

متن کامل

Texture Segmentation Based on the Oscillatory Feature

Texture segmentation is a typical difficult problem in image processing. This paper presents a new textural oscillatory feature based on image decomposition. The oscillatory feature together with other textural features based on the structure tensor and nonlinear diffusion constructs a 5 dimensional textural feature space. The last result can be obtained by segmenting the feature space using le...

متن کامل

Texture segmentation via a diffusion-segmentation scheme in the Gabor feature space

We address the problem of texture segmentation in the context of the Gabor feature space of images. Gabor filters which are tuned to different orientations, scales and frequencies are applied to textured images to create the Gabor feature space. We regard the scale, orientation and frequency for which maximum response of the Gabor filters was obtained. A two-dimensional Riemannian manifold of l...

متن کامل

Unsupervised Texture Image Segmentation Using MRFEM Framework

Texture image analysis is one of the most important working realms of image processing in medical sciences and industry. Up to present, different approaches have been proposed for segmentation of texture images. In this paper, we offered unsupervised texture image segmentation based on Markov Random Field (MRF) model. First, we used Gabor filter with different parameters’ (frequency, orientatio...

متن کامل

Unsupervised Texture Image Segmentation Using MRFEM Framework

Texture image analysis is one of the most important working realms of image processing in medical sciences and industry. Up to present, different approaches have been proposed for segmentation of texture images. In this paper, we offered unsupervised texture image segmentation based on Markov Random Field (MRF) model. First, we used Gabor filter with different parameters’ (frequency, orientatio...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2003