Authors: Mansuo Zhao, David Tien
Addresses: School of Accounting and Computer Science, Charles Sturt University, Bathurst, NSW, Australia. ' School of Accounting and Computer Science, Charles Sturt University, Bathurst, NSW, Australia
Abstract: An adaptive image segmentation method using Markov random field model for suburban aerial images is presented in this paper. The image is modelled as a collection of regions characterised by slowly moving averages and standard deviation. Decreasing sized windows are used to calculate the moving averages during the iteration process. A function based weighting parameter between the two components in the energy function is also used to improve the performance of unsupervised segmentation. A hierarchical implementation scheme is also introduced to reduce the computation load and increase the segmentation speed.
Keywords: aerial suburban images; hierarchical implementation; Markov random field model; adaptive image segmentation.
International Journal of Computer Applications in Technology, 2009 Vol.34 No.4, pp.235 - 240
Published online: 25 Mar 2009 *Full-text access for editors Access for subscribers Purchase this article Comment on this article