Multiphase Evolution and Variational Image Classification
نویسندگان
چکیده
This report presents a supervised classi cation model based on a variational approach. This model is devoted to nd an optimal partition compound of homogeneous classes with regular interfaces. We represent the regions of the image de ned by the classes and their interfaces by level set functions, and we de ne a functional whose minimum is an optimal partition. The coupled Partial Di erential Equations (PDE) related to the minimization of the functional are considered through a dynamical scheme. Given an initial interface set (zero level set), the di erent terms of the PDE's are governing the motion of interfaces such that, at convergence, we get an optimal partition as de ned above. Each interface is guided by internal forces (regularity of the interface), and external ones (data term, no vacuum, no regions overlapping). We conducted several experiments on both synthetic an real images. Key-words: Variational model, classi cation, labelling, level set formulation, active regions, active contours, multiphase, satellite images. Laboratoire J.A. dieudonné UMR 6621 CNRS, Université de Nice-Sophia Antipolis, 06108 Nice Cedex 2, FRANCE Evolution de Phases Multiples et Classi cation d'Images par modèle variationnel Résumé : Dans ce rapport, nous présentons un modèle de classi cation supervisée basé sur une approche variationnelle. Nous souhaitons obtenir une partition optimale de l'image constituée de classes homogènes séparées par des interfaces régulières. Pour cela, nous représentons les régions dé nies par les classes ainsi que leurs interfaces par des fonctions d'ensembles de niveaux. Nous dé nissons une fonctionnelle sur ces ensembles de niveaux dont le minimum est une partition optimale. Les Equations aux Dérivées Partielles (EDP) relatives à la minimisation de la fonctionnelle sont couplées et plongées dans une schéma dynamique. En xant un ensemble de niveaux initial, les di érents termes des EDP guident l'évolution des interfaces (ensembles de niveaux zéro) vers les frontières de la partition optimale, par le biais de forces internes (régularité de l'interface) et externes (attache aux données et pas de chevauchement des régions ni de vide dans la partition). Nous avons e ectué de nombreux tests sur des images synthétiques ainsi que sur des images réelles. Mots-clés : Modèle variationnel, classi cation d'images, formulation par ensembles de niveaux, régions actives, contours actifs, phases multiples, images satellitaires. Multiphase Evolution and Variational Image Classi cation 3
منابع مشابه
MULTIPHASE IMAGE SEGMENTATION BASED ON INTENSITY STATISTICS: MODELING AND APPLICATIONS By FUHUA CHEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MULTIPHASE IMAGE SEGMENTATION BASED ON INTENSITY STATISTICS: MODELING AND APPLICATIONS By Fuhua Chen August 2012 Chair: Yunmei Chen Major: Mathematics Image segmentation is the first stage of image processing and analysis. The result of ...
متن کاملA Continuous Labeling for Multiphase Graph Cut Image Partitioning
This study investigates a variational multiphase image segmentation method which combines the advantages of graph cut discrete optimization and multiphase piecewise constant image representation. The continuous region parameters serve both image representation and graph cut labeling. The algorithm iterates two consecutive steps: an original closed-form update of the region parameters and partit...
متن کاملNew Algorithm for Level Set Evolution without Re-initialization and Its Application to Variational Image Segmentation
Traditionally variational level set model for image segmentation is solved by using gradient descent method, which has low computational efficiency and needs complex re-initialization of level set functions as signed distance functions. In this paper, we first reformulate the variational model as a constrained optimization problem. Then we present an augmented Lagrangian projection method to pr...
متن کاملMultiple Motion and Occlusion Segmentation with a Multiphase Level Set Method
In this paper, we propose a new variational formulation for simultaneous multiple motion segmentation and occlusion detection in an image sequence. For the representation of segmented regions, we use the multiphase level set method proposed by Vese and Chan. This method allows an efficient representation of up to 2 regions with L level-set functions. Moreover, by construction, it enforces a dom...
متن کاملA Multiphase Image Segmentation Based on Fuzzy Membership Functions and L1-norm Fidelity
In this paper, we propose a variational multiphase image segmentation model based on fuzzy membership functions and L1-norm fidelity. Then we apply the alternating direction method of multipliers to solve an equivalent problem. All the subproblems can be solved efficiently. Specifically, we propose a fast method to calculate the fuzzy median. Experimental results and comparisons show that the L...
متن کاملذخیره 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 1999