文章基本信息

标题：A Novel Method For Generation Of Random Fields For Boundary Detection And Classification Of Images
本地全文：下载
作者：C. NagaRaju ; P.S.V.Srinivasa Rao ; G.Girija Rani 等
期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2008
卷号：8
期号：7
页码：331-337
出版社：International Journal of Computer Science and Network Security
摘要：Markov random field(MRF) theory has been widely applied to the challenging problem of Image Segmentation. Image segmentation is a task that classifies pixels of an Image using different labels so that the Image is partitioned into non-overlapping labeled regions. Image segmentation is one of the most difficult problems that researchers are facing because most of the real objects have complex shapes, boundaries and morphology, and true images are often corrupted by noise that cannot be ignored. To tackle the difficult problem of image segmentation, researchers have proposed a variety of methods. In this paper, a new texture segmentation method using compound MRFs is proposed, in which the label MRF and boundary MRF are coupled with gray level watershed method to help improve the segmentation performance. The boundary model is relatively general and does not need prior training on boundary patterns. Unlike some existing related work, the proposed method offers a more compact interaction between label, boundary MRFs with gray level water shed method. It is experimentally shown that proposed method can segment objects with complex boundaries and at the same time s able to work under noise corruption. The new method has been applied to medical image segmentation. Experiments on synthetic images and real clinical datasets show that the proposed method is able to produce more accurate segmentation results and satisfactorily keep the delicate boundary. It is also less sensitive to noise in both high and low signal-to-noise rati regions than some of the existing models in common use.
关键词：Boundary model, Markov random fields (MRFs), image segmentation, gray level watershed, Markovianity and frequency of gray levels.