University of Twente Student Theses


A Front-end Application for Markov Random Field-based Texture Image Segmentation

Huizenga, G.R. (2015) A Front-end Application for Markov Random Field-based Texture Image Segmentation.

[img] PDF
Abstract:Accurately segmenting texture images that are characterized by globally varying patterns that belong to the same class is a challenge, especially when the images are large or highly susceptible to noise. Such textures appear for example in radar images and in H&E stained pathology images. Markov Random Fields segmentation is a powerful technique that is able to take into account information of the surroundings of a pixel to infer the most likely class it belongs to. However, Markov Random Fields work on pixel level which makes it very computationally demanding, especially for large images or when fast processing times are required. Furthermore, working on pixel level means that it is hard to use accurate texture features during segmentation. This document aims to overcome these drawbacks of Markov Random Field segmentation by providing a front-end based on the concept of superpixels, the output of which can be used as input to the Markov RandomField segmentation algorithm.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Embedded Systems MSc (60331)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page