Unsupervised Change Detection in Satellite Images Using Principal Component Analysis and $k$-Means Clustering

Çelik, Turgay

doi:10.1109/lgrs.2009.2025059

Cited by 870 publications

(224 citation statements)

References 7 publications

Supporting

Mentioning

159

Contrasting

Unclassified

Order By: Relevance

“…To assess the effectiveness of the proposed method, we compared our method with other pixel-based CD methods including multi-resolution level set change-detection approach MLS [Bazi et al, 2010], Thresholding method [Ostu, 1979], PCA-Kmeans [Celik, 2009], Kernel method [Volpi et al, 2012], MRF [Bruzzone and Prieto, 2002] and traditional CRF based algorithms. In MLS, the parameter settings are the same to [Bazi et al, 2010].…”

Section: Resultsmentioning

confidence: 99%

“…X denote two co-registered multi-spectral remote sensing images acquired over the same geograhpical area at two different times. d X is the difference image generated from 1 X and 2 X by applying PCA technique [Celik, 2009].…”

Section: Proposed Change-detection Approachmentioning

confidence: 99%

“…Changes are then identified by analyzing the difference image. Celik [2009] presented a powerful difference image construction method based on principal component analysis (PCA). The difference image is partitioned into h h × non-overlapping blocks.…”

Section: Difference Imagementioning

confidence: 99%

“…Each pixel in the difference image is represented with an S -dimensional feature vector which is the projection of h h × difference image data onto the generated eigenvector space. We refer the reader to [Celik, 2009] for further details. Although Celik's method has increased the robustness against noise, we still face such limitations as noisy outputs when applying his algorithm in very high resolution images.…”

Section: S S Hmentioning

confidence: 99%

“…Celik [2010] presented a change-detection method by conducting probabilistic Bayesian inference with Expectation Maximization (EM) algorithm based parameter estimation to threshold DT-CWT data. Celik [2009] detected the changes using principal component analysis and K-means clustering method. Level set based methods [Bazi et al, 2010;Celik, 2011] have been introduced into changedetection applications recently.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Unsupervised change detection in high spatial resolution remote sensing images based on a conditional random field model

Cao

Zhou

2016

European Journal of Remote Sensing

View full text Add to dashboard Cite

In this paper, we propose a novel technique for unsupervised change detection in high spatial remote sensing images based on a conditional random field (CRF) model. The change-detection problem is formulated as a labeling issue to discriminate the changed class from the unchanged class in the difference image. CRF which employs the spatial property on both pixel's spectral data and labels have been widely used in many remote sensing applications. However, as there are a large number of model parameters to train, the CRF-based change-detection approach is time consuming and difficult to implement. The proposed method artfully uses memberships of Fuzzy C-means as unary potentials and defines pairwise potentials using a scaled squared Euclidean distance between neighboring pixels. This not only avoids training parameters but also helps improving the accuracy and the degree of automation. The experimental results obtained from three different remote sensing images demonstrate the accuracy and efficiency of our proposed method.

show abstract