Using Label Noise Robust Logistic Regression for Automated Updating of Topographic Geospatial Databases

Abstract

Supervised classification of remotely sensed images is a classical method to update topographic geospatial databases. The task requires training data in the form of image data with known class labels, whose generation is time-consuming. To avoid this problem one can use the labels from the outdated database for training. As some of these labels may be wrong due to changes in land cover, one has to use training techniques that can cope with wrong class labels in the training data. In this paper we adapt a label noise tolerant training technique to the problem of database updating. No labelled data other than the existing database are necessary. The resulting label image and transition matrix between the labels can help to update the database and to detect changes between the two time epochs. Our experiments are based on different test areas, using real images with simulated existing databases. Our results show that this method can indeed detect changes that would remain undetected if label noise were not considered in training.

GEOBIA at the Terapixel Scale: Toward Efficient Mapping of Small Woody Features from Heterogeneous VHR Scenes

“…It is known that this could introduce label noise or mislabeled samples in the training set due to several factors such as misregistration, out-dated maps and databases, etc. Thus, in a future work we would like to consider label noise robust classification methods [42,43] to improve the classification performance, and also to incorporate more valuable information and efficient techniques to further reduce the computation time while still increasing the classification accuracy.…”

3D change detection – Approaches and applications

“…-Civil -monitoring of structure, construction/mining progress, traffic and pedestrian tracking. (Grigillo et al, 2011;Nebiker et al, 2014), (Liu et al, 2003;Qin et al, 2015a) 3D model/map update: (Knudsen and Olsen, 2003;Li et al, 2008;Qin, 2014a), (Kim et al, 2013), (Maas et al, 2016) Disaster management: (Menderes et al, 2015), (Adams and Friedland, 2011), (Turker and Cetinkaya, 2005) , (Choi and Lee, 2011), (Gerke and Kerle, 2011) Landslides monitoring: (Martha et al, 2010), (Travelletti et al, 2012), (Ghuffar et al, 2013) Volcano eruption: (Hunter et al, 2003), (Baldi et al, 2005), (Vassilopoulou et al, 2002) Fault detection: (Copley et al, 2011), (Barisin et al, 2009).…”

“…Indeed a combinational use of both strategies (post-refinement and post-classification) might produce better results. When available, existing GIS data can be very helpful in both change refinement (for regularization) (Dini et al, 2012) and classification (for sample collection) (Maas et al, 2016) .…”

Unmanned aerial vehicle mapping for settlement upgrading

“…Random Forests are made up of a large number individual classification trees which are each trained individually using random feature and training sample subsets (Breiman, 2001). These methods are therefore particularly robust to errors in the training labels (Frenay and Verleysen, 2014;Maas, Rottensteiner and Heipke, 2016), but require a large number of training samples. (SVMs) are robust classifiers that are particularly suited to high dimensional feature spaces, have been proven to obtain high classification accuracies in remote sensing applications (Bruzzone and Persello, 2010), and can perform well with a limited number of training samples.…”

“…Other strategies have been developed to specifically combat label noise for remote sensing applications. For example, by modelling label noise by combining noise robust logistic regression and Conditional Random Fields (CRFs) for updating geospatial databases (Maas, Rottensteiner and Heipke, 2016); or by using the contextual information in a semi-supervised setting in order to assess the reliability of training samples and obtain a classification algorithm that is more robust to mislabeled training samples (Bruzzone and Persello, 2009).…”