Centre de documentation
scientifique

Dépouillements

An improved multi-task pointwise network for segmentation of building roofs in airborne laser scanning point clouds / Chaoquan Zhang in Photogrammetric record, vol 37 n° 179 (September 2022)  

Public [article]  inPhotogrammetric record > vol 37 n° 179 (September 2022) . - pp 260 - 284 Titre : An improved multi-task pointwise network for segmentation of building roofs in airborne laser scanning point clouds Type de document : Article/Communication Auteurs : Chaoquan Zhang, Auteur ; Hongchao Fan, Auteur Année de publication : 2022 Article en page(s) : pp 260 - 284 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Photogrammétrie [Termes IGN] analyse de groupement [Termes IGN] apprentissage profond [Termes IGN] classification barycentrique [Termes IGN] classification par réseau neuronal récurrent [Termes IGN] données d'entrainement (apprentissage automatique) [Termes IGN] fusion de données [Termes IGN] Norvège [Termes IGN] Ransac (algorithme) [Termes IGN] segmentation sémantique [Termes IGN] semis de points [Termes IGN] toit Résumé : (auteur) Roof plane segmentation is an essential step in the process of 3D building reconstruction from airborne laser scanning (ALS) point clouds. The existing approaches either rely on human intervention to select the appropriate input parameters for different data-sets or they are not automatic and efficient. To tackle these issues, an improved multi-task pointwise network is proposed to simultaneously segment instances (that is, individual roof planes) and semantics (that is, groups of roof planes with similar geometric shapes) in point clouds. PointNet++ is used as a backbone network to extract robust features in the first step. The features from semantics branch are then added to the instance branch to facilitate the learning of instance embeddings. After that, a feature fusion module is added to the semantics branch to acquire more discriminative features from the backbone network. To increase the accuracy of semantic predictions, fused semantic features of the points belonging to the same instance are aggregated together. Finally, a mean-shift clustering algorithm is employed on instance embeddings to produce the instance predictions. Furthermore, a new roof data-set (called RoofNTNU) is established by taking ALS point clouds as training data for automatic and more general segmentation. Experiments on the new roof data-set show that the method achieves promising segmentation results: the mean precision (mPrec) of 96.2% for the instance segmentation task and mean accuracy (mAcc) of 94.4% for the semantic segmentation task. Numéro de notice : A2022-936 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.1111/phor.12420 Date de publication en ligne : 13/07/2022 En ligne : https://doi.org/10.1111/phor.12420 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102682 [article]

Deflection of vertical effect on direct georeferencing in aerial mobile mapping systems: A case study in Sweden / Mohammad Bagherbandi in Photogrammetric record, vol 37 n° 179 (September 2022)  

Public [article]  inPhotogrammetric record > vol 37 n° 179 (September 2022) . - pp 285 - 305 Titre : Deflection of vertical effect on direct georeferencing in aerial mobile mapping systems: A case study in Sweden Type de document : Article/Communication Auteurs : Mohammad Bagherbandi, Auteur ; Arash Jouybari, Auteur ; Faramarz Nilfouroushan, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 285 - 305 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Photogrammétrie [Termes IGN] couplage GNSS-INS [Termes IGN] déviation de la verticale [Termes IGN] Earth Gravity Model 2008 [Termes IGN] ellipsoïde de référence [Termes IGN] géoréférencement direct [Termes IGN] photogrammétrie aérienne [Termes IGN] quasi-géoïde [Termes IGN] Suède [Termes IGN] système de numérisation mobile Résumé : (auteur) GNSS/INS applications are being developed, especially for direct georeferencing in airborne photogrammetry. Achieving accurately georeferenced products from the integration of GNSS and INS requires removing systematic errors in the mobile mapping systems. The INS sensor's uncertainty is decreasing; therefore, the influence of the deflection of verticals (DOV, the angle between the plumb line and normal to the ellipsoid) should be considered in the direct georeferencing. Otherwise, an error is imposed for calculating the exterior orientation parameters of the aerial images and aerial laser scanning. This study determines the DOV using the EGM2008 model and gravity data in Sweden. The impact of the DOVs on horizontal and vertical coordinates, considering different flight altitudes and camera field of view, is assessed. The results confirm that the calculated DOV components using the EGM2008 model are sufficiently accurate for aerial mapping system purposes except for mountainous areas because the topographic signal is not modelled correctly. Numéro de notice : A2022-937 Affiliation des auteurs : non IGN Thématique : IMAGERIE/POSITIONNEMENT Nature : Article DOI : 10.1111/phor.12421 Date de publication en ligne : 25/07/2022 En ligne : https://doi.org/10.1111/phor.12421 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102683 [article]