The true false ground truths: What interest?

The existence of a few unreliable ground truth (GT) data sets which are often used as reference by the remote sensing community for the assessment and comparison of classification results is really problematic and poses a number of questions. Two of these ground truth data sets can be cited: "Pavia University" and "Indian Pine". A rigorous analysis of spectral signatures of pixels in these images shows that some classes which are considered as homogeneous from the ground truth are clearly not, since the pixels which belong to the same classes have different spectral signatures, and probably do not belong to the same category. The persistence in using data sets from a biased ground truth does not allow objective comparisons between classification methods and does not contribute to providing explanation of physical phenomena that images are supposed to reflect. In this communication, we present a fine and complete analysis of the spectral signatures of pixels within each class for the two ground truth data sets mentioned above. The metrics used show some incoherence and inaccuracy of these data which wrongly serve as references in several classification comparative studies.

[1]  Jon Atli Benediktsson,et al.  Segmentation and Classification of Hyperspectral Data using Watershed , 2008, IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium.

[2]  Qian Du,et al.  Classification of hyperspectral urban data using adaptive simultaneous orthogonal matching pursuit , 2014 .

[3]  Kuo-Lung Wu,et al.  Analysis of parameter selections for fuzzy c-means , 2012, Pattern Recognit..

[4]  Delbert Dueck,et al.  Clustering by Passing Messages Between Data Points , 2007, Science.

[5]  Paul Claval,et al.  Le rôle du terrain en géographie. Des épistémologies de la curiosité à celles du désir , 2013 .

[6]  Robert M. Gray,et al.  An Algorithm for Vector Quantizer Design , 1980, IEEE Trans. Commun..

[7]  Peng Gao,et al.  Application of fuzzy c-means clustering in data analysis of metabolomics. , 2009, Analytical chemistry.

[8]  Jon Atli Benediktsson,et al.  Recent Advances in Techniques for Hyperspectral Image Processing , 2009 .

[9]  Lorenzo Bruzzone,et al.  Classification of hyperspectral remote sensing images with support vector machines , 2004, IEEE Transactions on Geoscience and Remote Sensing.

[10]  Claude Cariou,et al.  Unsupervised Nearest Neighbors Clustering With Application to Hyperspectral Images , 2015, IEEE Journal of Selected Topics in Signal Processing.

[11]  Qingquan Li,et al.  Spectral–Spatial Hyperspectral Image Classification Using $\ell_{1/2}$ Regularized Low-Rank Representation and Sparse Representation-Based Graph Cuts , 2015, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[12]  Liangpei Zhang,et al.  An Adaptive Artificial Immune Network for Supervised Classification of Multi-/Hyperspectral Remote Sensing Imagery , 2012, IEEE Transactions on Geoscience and Remote Sensing.

[13]  Behnaz Bigdeli,et al.  A Multiple SVM System for Classification of Hyperspectral Remote Sensing Data , 2013, Journal of the Indian Society of Remote Sensing.

[14]  Xin Huang,et al.  A comparative study of spatial approaches for urban mapping using hyperspectral ROSIS images over Pavia City, northern Italy , 2009 .

[15]  Nikolaos Doulamis,et al.  Deep supervised learning for hyperspectral data classification through convolutional neural networks , 2015, 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).

[16]  Jon Atli Benediktsson,et al.  SVM- and MRF-Based Method for Accurate Classification of Hyperspectral Images , 2010, IEEE Geoscience and Remote Sensing Letters.

[17]  J. MacQueen.  Some methods for classification and analysis of multivariate observations , 1967 .

[18]  Arnaldo de Albuquerque Araújo,et al.  Unsupervised Hyperspectral Band Selection Based on Spectral Rhythm Analysis , 2014, 2014 27th SIBGRAPI Conference on Graphics, Patterns and Images.

[19]  Claude Cariou,et al.  Pixel classification of large-size hyperspectral images by affinity propagation , 2014 .

[20]  A. Araújo,et al.  Feature Selection for Classification of Remote Sensed Hyperspectral Images : A Filter approach using Genetic Algorithm and Cluster Validity , 2012 .

[21]  K. Chehdi,et al.  Unsupervised and stable LBG algorithm for data classification: application to aerial multicomponent images , 2015, SPIE Remote Sensing.

[22]  Jon Atli Benediktsson,et al.  Generalized Composite Kernel Framework for Hyperspectral Image Classification , 2013, IEEE Transactions on Geoscience and Remote Sensing.