DocumentCode :
937859
Title :
Morse operators for digital planar surfaces and their application to image segmentation
Author :
Nonato, Luis Gustavo ; Filho, Antônio Castelo ; Minghim, Rosane ; Batista, João
Author_Institution :
Dept. de Ciencias da Computacao e Estatistica, Univ. de Sao Paulo, Sao Carlos, Brazil
Volume :
13
Issue :
2
fYear :
2004
Firstpage :
216
Lastpage :
227
Abstract :
This paper introduces the concept of digital planar surfaces and corresponding Morse operators. These operators offer a novel and powerful method for construction and de-construction of such surfaces in a way that global topological control of the resulting object is always maintained. In that respect, this paper offers a complete pixel characterization tool. Image handling is a natural application for such approach. We present a novel fast algorithm for image segmentation using Morse operators for digital planar surfaces. It classifies as a region growing technique with added topological control and is extremely useful for applications that need proper object description. Results from real data are stimulating, and show that the segmentation algorithm compares very well with other methods. The topological approach also forms a base for future expansion to applications such as volume segmentation.
Keywords :
computation theory; image segmentation; medical image processing; Morse operators; Morse theory; computational topology; digital planar surfaces; global topological control; image handling; image segmentation; medical imaging; object description; pixel characterization tool; region growing technique; Application software; Biomedical imaging; Computational modeling; Digital images; Image segmentation; Robustness; Solid modeling; Topology; Algorithms; Image Enhancement; Image Interpretation, Computer-Assisted; Imaging, Three-Dimensional; Information Storage and Retrieval; Pattern Recognition, Automated; Reproducibility of Results; Sensitivity and Specificity; Signal Processing, Computer-Assisted;
fLanguage :
English
Journal_Title :
Image Processing, IEEE Transactions on
Publisher :
ieee
ISSN :
1057-7149
Type :
jour
DOI :
10.1109/TIP.2003.819908
Filename :
1278336
Link To Document :
بازگشت