Centre de documentation
scientifique

Invariant structure representation for remote sensing object detection based on graph modeling / Zicong Zhu in IEEE Transactions on geoscience and remote sensing, vol 60 n° 6 (June 2022)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 60 n° 6 (June 2022) . - n° 5625217 Titre : Invariant structure representation for remote sensing object detection based on graph modeling Type de document : Article/Communication Auteurs : Zicong Zhu, Auteur ; Xian Sun, Auteur ; Wenhui Diao, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : n° 5625217 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] détection d'objet [Termes IGN] données d'entrainement sans étiquette [Termes IGN] extraction de traits caractéristiques [Termes IGN] filtrage numérique d'image [Termes IGN] granularité d'image [Termes IGN] graphe [Termes IGN] invariant Résumé : (auteur) Due to the characteristics of vertical orthophoto imaging, the apparent structural features of the object in the remote sensing (RS) image are relatively stable, such as the cross-shaped structure of the aircraft and the rectangular structure of the vehicle. Compared with the traditional visual features, using these features is conducive to improving the accuracy of object detection. However, there are few studies on such characteristics. In this article, we systematically study the invariant structural features of remote sensing objects and propose a graph focusing aggregation network (GFA-Net) to represent the structural features of remote sensing objects. Among them, in view of the problem that traditional convolutional neural networks (CNNs) are sensitive to the changes in rotation, scale, and other factors, which makes it difficult to extract structural features, we propose the graph focusing process (GFP) based on the idea of graph convolution. Analysis and experiments show that graph structure has significant advantages over Euclidean feature space under CNN in expressing such structural features. In order to realize the end-to-end efficient training of the above model, we design a graph aggregation network (GAN) to update the weight of nodes. We verify the effectiveness of our method on the proposed multitask datasets aircraft component segmentation dataset (ACSD) and the large-scale Fine-grAined object recognItion in high-Resolution RS imagery (FAIR1M). Experiments conducted on the object detection datasets of large-scale Dataset for Object deTection in Aerial images (DOTA) and HRSC2016 prove that the proposed method is superior to the current state-of-the-art (SOTA) method. Numéro de notice : A2022-560 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2022.3181686 Date de publication en ligne : 09/06/2022 En ligne : https://doi.org/10.1109/TGRS.2022.3181686 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101186 [article]

VD-LAB: A view-decoupled network with local-global aggregation bridge for airborne laser scanning point cloud classification / Jihao Li in ISPRS Journal of photogrammetry and remote sensing, vol 186 (April 2022)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 186 (April 2022) . - pp 19 - 33 Titre : VD-LAB: A view-decoupled network with local-global aggregation bridge for airborne laser scanning point cloud classification Type de document : Article/Communication Auteurs : Jihao Li, Auteur ; Martin Weinmann, Auteur ; Xian Sun, Auteur ; et al., Auteur Année de publication : 2022 Article en page(s) : pp 19 - 33 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique [Termes IGN] agrégation de détails [Termes IGN] apprentissage profond [Termes IGN] précision de la classification [Termes IGN] qualité du modèle [Termes IGN] semis de points [Termes IGN] télémétrie laser aéroporté Résumé : (Auteur) Airborne Laser Scanning (ALS) point cloud classification is a valuable and practical task in the fields of photogrammetry and remote sensing. It takes an extremely important role in many applications of surveying, monitoring, planning, production and living. Recently, driven by the wave of deep learning, the classification of ALS point clouds has also been gradually shifting from traditional feature design to careful deep network architecture construction. Although significant progress has been achieved by leveraging deep learning technology, there are still some matters demanding prompt solution: (1) the coupling phenomenon of high-level semantic features from multiple field-of-views; (2) information propagation without aggregated local–global features in different levels of symmetrical structure; (3) quite serious class-imbalanced distribution problems in large-scale ALS point clouds. In this paper, to tackle these matters, we propose a novel View-Decoupled Network with Local–global Aggregation Bridge (VD-LAB) model. More concretely, a View-Decoupled (VD) grouping method is set at the deepest layer of the network. Then, we establish a Local–global Aggregation Bridge (LAB) between down-sampling path and up-sampling path of the same level. After that, a Self-Amelioration (SA) loss is taken as the optimization objective to train the whole model in an end-to-end manner. Extensive experiments on four challenging ALS point cloud datasets (LASDU, US3D, ISPRS 3D and GML) demonstrate that our VD-LAB is able to achieve state-of-the-art performance in terms of Overall Accuracy (OA) and mean -score (e.g., reaching 88.01% and 78.42% for LASDU dataset, respectively) with very few model parameters and it possesses a strong generalization capability. In addition, the visualization of achieved results also reveals more satisfactory classification for some categories, such as Water in the US3D dataset and Powerline in the ISPRS 3D dataset. Ultimately, the effect of each module in VD-LAB is assessed in detailed ablation analyses as well. Numéro de notice : A2022-067 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2022.01.012 Date de publication en ligne : 10/02/2022 En ligne : https://doi.org/10.1016/j.isprsjprs.2022.01.012 Format de la ressource électronique : URL Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=99789 [article]

Exemplaires(3)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
081-2022041	SL	Revue	Centre de documentation	Revues en salle	Disponible
081-2022043	DEP-RECP	Revue	LASTIG	Dépôt en unité	Exclu du prêt
081-2022042	DEP-RECF	Revue	Nancy	Dépôt en unité	Exclu du prêt

PBNet: Part-based convolutional neural network for complex composite object detection in remote sensing imagery / Xian Sun in ISPRS Journal of photogrammetry and remote sensing, vol 173 (March 2021)  

Public [article]  inISPRS Journal of photogrammetry and remote sensing > vol 173 (March 2021) . - pp 50 - 65 Titre : PBNet: Part-based convolutional neural network for complex composite object detection in remote sensing imagery Type de document : Article/Communication Auteurs : Xian Sun, Auteur ; Peijin Wang, Auteur ; Cheng Wang, Auteur ; et al., Auteur Année de publication : 2021 Article en page(s) : pp 50 - 65 Note générale : bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection [Termes IGN] apprentissage profond [Termes IGN] Chine [Termes IGN] classification par réseau neuronal convolutif [Termes IGN] détection d'objet [Termes IGN] objet géographique complexe [Termes IGN] prise en compte du contexte [Termes IGN] rectangle englobant minimum Résumé : (auteur) In recent years, deep learning-based algorithms have brought great improvements to rigid object detection. In addition to rigid objects, remote sensing images also contain many complex composite objects, such as sewage treatment plants, golf courses, and airports, which have neither a fixed shape nor a fixed size. In this paper, we validate through experiments that the results of existing methods in detecting composite objects are not satisfying enough. Therefore, we propose a unified part-based convolutional neural network (PBNet), which is specifically designed for composite object detection in remote sensing imagery. PBNet treats a composite object as a group of parts and incorporates part information into context information to improve composite object detection. Correct part information can guide the prediction of a composite object, thus solving the problems caused by various shapes and sizes. To generate accurate part information, we design a part localization module to learn the classification and localization of part points using bounding box annotation only. A context refinement module is designed to generate more discriminative features by aggregating local context information and global context information, which enhances the learning of part information and improve the ability of feature representation. We selected three typical categories of composite objects from a public dataset to conduct experiments to verify the detection performance and generalization ability of our method. Meanwhile, we build a more challenging dataset about a typical kind of complex composite objects, i.e., sewage treatment plants. It refers to the relevant information from authorities and experts. This dataset contains sewage treatment plants in seven cities in the Yangtze valley, covering a wide range of regions. Comprehensive experiments on two datasets show that PBNet surpasses the existing detection algorithms and achieves state-of-the-art accuracy. Numéro de notice : A2021-105 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2020.12.015 Date de publication en ligne : 16/01/2021 En ligne : https://doi.org/10.1016/j.isprsjprs.2020.12.015 Format de la ressource électronique : url article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=96891 [article]

Exemplaires(3)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
081-2021031	SL	Revue	Centre de documentation	Revues en salle	Disponible
081-2021033	DEP-RECP	Revue	LASTIG	Dépôt en unité	Exclu du prêt
081-2021032	DEP-RECF	Revue	Nancy	Dépôt en unité	Exclu du prêt