The watershed algorithm from mathematical morphology is powerful for segmentation and often produces more stable segmentation results, including continuous segmentation boundaries. However, direct application of the watershed segmentation algorithm generally leads to oversegmentation due to noise and other local irregularities of the gradient. Oversegmentation can be serious enough to render the result of the alorithm virtually useless. An approach used to control oversegmentation is based on the concept of markers. A marker is a connected component belonging to an image. In this paper, considering that there are two small but very brighter regions associated with the chimneys of ship, the topological properties of IR ship target can be used to restrict the algorithm. So, we have internal markers, associated with IR ship target, and external markers, associated with the background. Experimental results show that, when the original watershed segmentation has noisy boundaries or wrong limbs attached to the object of interest, the proposed method overcomes those drawbacks and yields a better segmentation.

Date Published: 10 January 2005
PDF: 4 pages
Proc. SPIE 5640, Infrared Components and Their Applications, (10 January 2005); doi: 10.1117/12.576118

Published in SPIE Proceedings Vol. 5640:
Infrared Components and Their Applications
Haimei Gong; Yi Cai; Jean-Pierre Chatard, Editor(s)