A low resolution ETM+ and higher resolution IKONOS images obtained in 2000 - 2003, respectively, are compared using post-classification comparison algorithm to detect changes due to a new major highway construction initiated in 2000. The spatial resolution difference between the input images, which may lead to wrong registration, thereby, wrong post-classification comparison results, are minimized by fusing the 30 meter resolution ETM+ multispectral bands with its 15 meter resolution panchromatic band using à trous wavelet transform image fusion method. The IKONOS image has color change on the sea surface especially at the river mouth and its vicinity because of mud and sediments carried by Degirmendere Creek. Therefore, maximum Likelihood, Spectral Angular Mapping, Fisher Linear Likelihood, and ECHO classifiers are used for image classification. The best results are obtained from the ECHO classifier since it considers both spectral and spatial variations in the input images. The results show that the land fill on costal zone due to new high way construction is detected successfully with ECHO classifier. It is also seen that improving the spatial resolution of the ETM+ via image fusion minimizes the impact of misclassification on final change image generated by post-classification comparison.
Key words: Change detection, image fusion, à trous wavelet transform, ETM+, IKONOS, ECHO classifier.
Copyright © 2020 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0