Centre de documentation
scientifique

Dépouillements

Vehicle detection in aerial images / Michael Ying Yang in Photogrammetric Engineering & Remote Sensing, PERS, vol 85 n° 4 (avril 2019)  

Public [article]  inPhotogrammetric Engineering & Remote Sensing, PERS > vol 85 n° 4 (avril 2019) . - pp 297 - 304 Titre : Vehicle detection in aerial images Type de document : Article/Communication Auteurs : Michael Ying Yang, Auteur ; Wentong Liao, Auteur ; Xinbo Li, Auteur ; et al., Auteur Année de publication : 2019 Article en page(s) : pp 297 - 304 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] apprentissage profond [Termes IGN] compréhension de l'image [Termes IGN] détection d'objet [Termes IGN] entropie [Termes IGN] image aérienne [Termes IGN] orthoimage [Termes IGN] précision de la classification [Termes IGN] qualité d'image [Termes IGN] réseau neuronal convolutif [Termes IGN] véhicule automobile Résumé : (Auteur) The detection of vehicles in aerial images is widely applied in many applications. Comparing with object detection in the ground view images, vehicle detection in aerial images remains a challenging problem because of small vehicle size and the complex background. In this paper, we propose a novel double focal loss convolutional neural network (DFL-CNN) framework. In the proposed framework, the skip connection is used in the CNN structure to enhance the feature learning. Also, the focal loss function is used to substitute for conventional cross entropy loss function in both of the region proposal network (RPN) and the final classifier. We further introduce the first large-scale vehicle detection dataset ITCVD with ground truth annotations for all the vehicles in the scene. We demonstrate the performance of our model on the existing benchmark German Aerospace Center (DLR) 3K dataset as well as the ITCVD dataset. The experimental results show that our DFL-CNN outperforms the baselines on vehicle detection. Numéro de notice : A2019-163 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.14358/PERS.85.4.297 Date de publication en ligne : 01/04/2019 En ligne : https://doi.org/10.14358/PERS.85.4.297 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=92568 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
105-2019041	SL	Revue	Centre de documentation	Revues en salle	Disponible

Integrated relative orientation based on point and line features via Plücker coordinates / Qinghong Sheng in Photogrammetric Engineering & Remote Sensing, PERS, vol 85 n° 4 (avril 2019)  

Public [article]  inPhotogrammetric Engineering & Remote Sensing, PERS > vol 85 n° 4 (avril 2019) . - pp 305 - 311 Titre : Integrated relative orientation based on point and line features via Plücker coordinates Type de document : Article/Communication Auteurs : Qinghong Sheng, Auteur ; Rui Yang, Auteur ; Hui Xiao, Auteur Année de publication : 2019 Article en page(s) : pp 305 - 311 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] méthode des moindres carrés [Termes IGN] orientation relative [Termes IGN] transformation de coordonnées Résumé : (Auteur) Relative orientation based on point and line features can significantly reduce the problems caused by single flight strip bending and strengthen geometric stability. Nevertheless, the dimensional difference of points and lines and the inconsistency of their methods of representation result in low calculation efficiency. In this paper, it is proposed that three-dimensional points and four-dimensional lines are uniformly represented by Plücker coordinates to achieve integrated coordinate transformation. According to the geometric conditions that lines intersect lines and planes intersect planes, a relative orientation model based on pointline via Plücker coordinates (P-LPRO) is established, and the error equation is linearized by the least-squares method. Experimental results show that the integrated adjustment solution of P-LPRO is more accurate than the traditional method and can achieve faster convergence speed. Numéro de notice : A2019-164 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.14358/PERS.85.4.305 Date de publication en ligne : 01/04/2019 En ligne : https://doi.org/10.14358/PERS.85.4.305 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=92569 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
105-2019041	SL	Revue	Centre de documentation	Revues en salle	Disponible