Image segmentation using texture information and graph cut on images modelled by means of hierachical structures

• Grant number: 13/00575-3
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: April 01, 2013
• End date: March 31, 2014
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Marco Antonio Garcia de Carvalho
• Grantee: Kauê Tartarotti Nepomuceno Duarte
• Host Institution: Faculdade de Tecnologia (FT). Universidade Estadual de Campinas (UNICAMP). Limeira , SP, Brazil

Abstract

Image segmentation is an important task in many areas of image processing, computer vision, and pattern recognition. Subdivide an image into its constituent parts may, oftentimes, be dependent on the context where the segmentation is applied. There are many techniques of image segmentation, being the use of graphs as structures to represent images quite common. Thus, the study of graph cut, or graph partitioning, corresponds to the process of image segmentation. A recent method used to partition graphs comes from Spectral Graph Theory, in which an analysis is made of the eigenvectors from an array representing a graph. The goal of this project is to perform image segmentation using the graph normalized cut technique and which uses concepts of Spectral Graph Theory. It is intended that the modeling graph takes into account aspects related to pixel information and primitive regions, being used the feature texture in particular. The images are represented by different types of graphs and these will have their effectiveness and applicability compared with respect to the segmentation task. This project provides the following contributions: investigate and explore the digital image segmentation task through graph cut and Spectral Graph Theory using texture features in image representations and consolidate measures of performance evaluation and comparison of the results of segmentation obtained by different strategies of graph cut. This project will be developed in the Computing Visual Laboratory (IMAGELab) of the School of Technology at UNICAMP. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

DUARTE, KAUE T. N.; DE CARVALHO, MARCO A. G.; MARTINS, PAULO S.; BLANCTALON, J; PENNE, R; PHILIPS, W; POPESCU, D; SCHEUNDERS, P. Adding GLCM Texture Analysis to a Combined Watershed Transform and Graph Cut Model for Image Segmentation. ADVANCED CONCEPTS FOR INTELLIGENT VISION SYSTEMS (ACIVS 2017), v. 10617, p. 12-pg., 2017-01-01. (13/00575-3)