Image Segmentation with Topological Maps and Inter-pixel Representation

✍ Scribed by Jean-Pierre Braquelaire; Luc Brun

• Publisher: Elsevier Science
• Year: 1998
• Tongue: English
• Weight: 945 KB
• Volume: 9
• Category: Article
• ISSN: 1047-3203
• DOI: 10.1006/jvci.1998.0374

✦ Synopsis

In this paper we present a data structure that improves region segmentation of 2D images. This data structure provides an

• The mixed approach which leads to split and merge efficient access to both geometric features such as the set of algorithms [4,13, 20, 21,44], introduced by Horowitz and pixels of a region and topological features like the frontier of Pavlidis [29, 30] and combining splitting and merging.

a region, the neighbors of a region or the set of regions included

All these methods partition the original image by recurin one region. It allows us to combine different segmentation algorithms without restriction. Interactive refinement or merge sively splitting and/or merging its regions. In the following, of regions can also be performed efficiently.