Constellations and the unsupervised learning of graphs

In this paper, we propose a novel method for the unsupervised clustering of graphs in the context of the constellation approach to object recognition. Such method is an EM central clustering algorithm which builds prototypical graphs on the basis of fast matching with graph transformations. Our experiments, both with random graphs and in realistic situations (visual localization), show that our prototypes improve the set median graphs and also the prototypes derived from our previous incremental method. We also discuss how the method scales with a growing number of images.

Imaging Land Subsidence Induced by Groundwater Extraction in Beijing (China) Using Satellite Radar Interferometry

Beijing is one of the most water-stressed cities in the world. Due to over-exploitation of groundwater, the Beijing region has been suffering from land subsidence since 1935. In this study, the Small Baseline InSAR technique has been employed to process Envisat ASAR images acquired between 2003 and 2010 and TerraSAR-X stripmap images collected from 2010 to 2011 to investigate land subsidence in the Beijing region. The maximum subsidence is seen in the eastern part of Beijing with a rate greater than 100 mm/year. Comparisons between InSAR and GPS derived subsidence rates show an RMS difference of 2.94 mm/year with a mean of 2.41 ± 1.84 mm/year. In addition, a high correlation was observed between InSAR subsidence rate maps derived from two different datasets (i.e., Envisat and TerraSAR-X). These demonstrate once again that InSAR is a powerful tool for monitoring land subsidence. InSAR derived subsidence rate maps have allowed for a comprehensive spatio-temporal analysis to identify the main triggering factors of land subsidence. Some interesting relationships in terms of land subsidence were found with groundwater level, active faults, accumulated soft soil thickness and different aquifer types. Furthermore, a relationship with the distances to pumping wells was also recognized in this work.