Change detection of multi-temporal remote sensing data based on image fusion and fuzzy clustering

We present a novel method based on image fusion and fuzzy clustering for change detection in remote sensing images. Firstly, the difference images by differencing operator and log-ratio operator are decomposed using Undecimated Discrete Wavelet transform (UDWT). Then, wavelet coefficients are fused to obtain multiscale feature vectors. By taking into account the spatial-neighborhood information, a change detection algorithm based fuzzy local-information C-means (FLICM) clustering is developed to obtain final detection results. Experiment confirms that the proposed approach does better than the differencing and log-ratio operators.