Bootstrapping CNNs for building segmentation in aerial imagery with depth

C. Sebastian, Bas Boom, Thijs M.B. van Lankveld, P.H.N. de With

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

4 Downloads (Pure)

Samenvatting

Detection of buildings and other objects from aerial images has various applications in urban planning and map making. For human experts, manual extraction of objects from aerial imagery is a time-consuming and expensive task. However, aerial images are often prone to varying lighting conditions, shadows and occlusions.Aerial images also differ quite significantly from region to region. In order to offer both robust and reliable building detection, we use Convolutional Neural Networks (CNN) to account for these variations. We train detection from RGB-D images to obtain a segmented mask by employing the CNN architecture DenseNet. To improve the performance of the model, we apply the statistical re-sampling technique called Bootstrapping. Through Bootstrapping, we demonstrate that more informative examples are retained that improves the results. Finally, the proposed methodology outperforms the non-bootstrapped DenseNet by utilizing only one-sixth of the original training data and it obtains a precision-recall break-even point at 95.10% score on our aerial imagery dataset.
Originele taal-2Engels
Titel2017 Symposium on Information Theory and Signal Processing in the Benelux
Aantal pagina's8
StatusGepubliceerd - mei 2017
Evenement2017 Symposium on Information Theory and Signal Processing in the Benelux (SITB 2017) - Delft University of Technology, Delft, Nederland
Duur: 11 mei 201712 mei 2017
http://cas.tudelft.nl/sitb2017/

Congres

Congres2017 Symposium on Information Theory and Signal Processing in the Benelux (SITB 2017)
Verkorte titelSITB 2017
Land/RegioNederland
StadDelft
Periode11/05/1712/05/17
Internet adres

Vingerafdruk

Duik in de onderzoeksthema's van 'Bootstrapping CNNs for building segmentation in aerial imagery with depth'. Samen vormen ze een unieke vingerafdruk.

Citeer dit