Multivariate alteration detection and ChiMerge thresholding method for change detection in bitemporal and multispectral images

This paper deals with an unsupervised approach for land change detection and extraction using bitemporal and multispectral remotely sensed images. It is a statistical approach based on multivariate alteration detection (MAD) transformation combined with a new ChiMerge thresholding method. As opposed to most other multivariate change detection schemes the MAD technique is invariant to affine transformations of the originally measured variables. Therefore, it is insensitive to linear differences in atmospheric conditions or sensor calibrations of multitemporal acquisitions. ChiMerge thresholding method is applied to automatically extract change detected by means of MAD analysis. A case study with SPOT-HRV multispectral images before and after a flood event occurred in November 2000 over Gloucester (UK) region shows the power of the proposed MAD/ChiMerge change detection scheme compared to the classical principal component analysis of simple difference combined to the same thresholding method PCA/ChiMerge. A comparison of extracted change objects is also carried out according to the ground truth of the study zone.

[1]  Zbigniew Bochenek,et al.  Change Detection Algorithm for the Production of Land Cover Change Maps over the European Union Countries , 2014, Remote. Sens..

[2]  John R. Jensen,et al.  Introductory Digital Image Processing: A Remote Sensing Perspective , 1986 .

[3]  Randy Kerber,et al.  ChiMerge: Discretization of Numeric Attributes , 1992, AAAI.

[4]  H. Hotelling.  Relations Between Two Sets of Variates , 1936 .

[5]  Du Q. Huynh,et al.  Unsupervised change detection based on robust chi-squared transform for bitemporal remotely sensed images , 2014 .

[6]  Hui Lin,et al.  Remote sensing change detection based on canonical correlation analysis and contextual bayes decision , 2007 .

[7]  Qian Du,et al.  Multi-Modal Change Detection, Application to the Detection of Flooded Areas: Outcome of the 2009–2010 Data Fusion Contest , 2012, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[8]  Allan Aasbjerg Nielsen,et al.  Change detection by the MAD method in hyperspectral image data , 2003 .

[9]  Allan Aasbjerg Nielsen,et al.  Multi-Channel Remote Sensing Data and Orthogonal Transformations for Change Detection , 1999 .

[10]  Knut Conradsen,et al.  Multivariate Alteration Detection (MAD) and MAF Postprocessing in Multispectral, Bitemporal Image Data: New Approaches to Change Detection Studies , 1998 .

[11]  Allan Aasbjerg Nielsen,et al.  The Regularized Iteratively Reweighted MAD Method for Change Detection in Multi- and Hyperspectral Data , 2007, IEEE Transactions on Image Processing.

[12]  Marvin E. Bauer,et al.  Processing of multitemporal Landsat TM imagery to optimize extraction of forest cover change features , 1994, IEEE Trans. Geosci. Remote. Sens..

[13]  Véronique Prinet,et al.  Multi-block PCA method for image change detection , 2003, 12th International Conference on Image Analysis and Processing, 2003.Proceedings..

[14]  H. Hotelling.  Analysis of a complex of statistical variables into principal components. , 1933 .

[15]  John A. Richards,et al.  Thematic mapping from multitemporal image data using the principal components transformation , 1984 .