Descripteur
Termes IGN > informatique > intelligence artificielle > vision par ordinateur
vision par ordinateurVoir aussi |
Documents disponibles dans cette catégorie (92)
Ajouter le résultat dans votre panier
Visionner les documents numériques
Affiner la recherche Interroger des sources externes
Etendre la recherche sur niveau(x) vers le bas
3D stem modelling in tropical forest: towards improved biomass and biomass change estimates / Sébastien Bauwens (2022)
Titre : 3D stem modelling in tropical forest: towards improved biomass and biomass change estimates Type de document : Thèse/HDR Auteurs : Sébastien Bauwens, Auteur Editeur : Gembloux [Belgique] : Université de Liège - Gembloux Agro-Bio Tech Année de publication : 2022 Importance : 146 p. Format : 21 x 30 cm Note générale : Bibliographie
Dissertation originale présentée en vue de l'obtention du grade de Docteur en Sciences Agronomiques et Ingénierie BiologiqueLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques
[Termes IGN] analyse comparative
[Termes IGN] biomasse aérienne
[Termes IGN] cartographie et localisation simultanées
[Termes IGN] Congo
[Termes IGN] diamètre à hauteur de poitrine
[Termes IGN] dioxyde de carbone
[Termes IGN] données lidar
[Termes IGN] écosystème forestier
[Termes IGN] forêt tropicale
[Termes IGN] hauteur des arbres
[Termes IGN] inventaire forestier (techniques et méthodes)
[Termes IGN] lidar mobile
[Termes IGN] modèle de croissance végétale
[Termes IGN] modèle numérique de terrain
[Termes IGN] placette d'échantillonnage
[Termes IGN] puits de carbone
[Termes IGN] semis de points
[Termes IGN] stéréoscopie
[Termes IGN] structure-from-motion
[Termes IGN] télémétrie laser terrestreRésumé : (auteur) Tropical forests are the main contributors of CO2 emissions between the biosphere and the atmosphere in the land use sector. The deforestation and degradation of these forests are the main sources of emissions from this sector, which accounts for 15% of the world's CO2 emissions. The monitoring of CO2 emissions and removals from tropical forests requires fine measurements of their trees. These measurements are then used as inputs in allometric model to predict the tree aboveground biomass and thus indirectly their equivalent in CO2. However, a significant proportion of trees in tropical forests show morphological singularities on the stem such as buttresses or other irregularities. The height (HPOM) of the diameter measured (DPOM) is therefore commonly raised above the buttresses to reach a circular part of the stem. The standard of measuring the diameter at breast height (DBH) is then lost. In this context, this thesis aims to improve the monitoring of tropical trees with stem irregularities by using recent three-dimensional (3D) measurement tools and developing a model-based approach to harmonize height measurements of the diameterdo. First, we evaluated the potential of the close-range terrestrial photogrammetric approach (CRTP) to measure irregular shaped stems. The advantage of this 3D approach is its low cost and ease of implementation as it only requires a camera and targets. Following the convincing results of this approach, we studied the quality of the allometric relationship between variables extracted from the stem cross-section at 1.3 m height and above-ground biomass. We found that the equivalent diameter of the basal area at 1.3 m height (DBH') correlates better with aboveground tree biomass and thus its carbon content than does diameter above buttress (DPOM). Therefore, harmonization of HPOM to 1.3 m height should be further studied to improve biomass estimates. Secondly, we investigated the potential of a hand-held mobile lidar scanner (HMLS) to measure in 3D not only one tree at a time but many trees from forest plots with a 15 m radius in Belgian temperate forest. To assess the HMLS, we compared it to 3D measurements made with a more commonly used static terrestrial laser scanning (TLS) and with conventional forest inventory diameter and position measurements. The HMLS has a better 3D spatial coverage of the stems than the TLS and the precision of the stem diameter measurements is also better with the HMLS. Setting up the plot and scanning it from five locations with the TLS takes three times longer than scanning with HMLS. This pioneering work shows us the potential of using HMLS in tropical forests through its speed of execution and its important spatial coverage at the stem level, an important issue for irregular shaped tree stems. Thirdly, we developed and assessed a model-based approach for harmonizing HPOM to correct the bias induced by irregular stems in the aboveground biomass estimates of forest inventory plots. Following the estimation of DBH' using a taper model proposed in our study, we find that conventional aboveground biomass estimates (i.e. with only DPOM), compared to estimates made with DBH', show an increasing divergence with the increase of irregular stems proportion within plots and going up to -15% in our study. These results show the importance of considering HPOM when estimating aboveground biomass in tropical forests, especially in forests with many irregular stems. Estimates of the evolution of plot above-ground biomass over time should also be revised to better consider the biomass growth of irregular shaped tree stems, which has been underestimated until now. Finally, based on the results of this research, we summarize the 3D measurement tools currently available and describe their advantages and disadvantages in the case of irregular stems. Based on available human and technical resources, we also give recommendations on the harmonization method to use in permanent sampling plots to correct the bias induced by irregular stems. Improved monitoring of these tropical trees may provide a better understanding of some of the residual, i.e. unexplained, terrestrial ecosystem CO2 sink currently noted in IPCC reports. Note de contenu : 1- General introduction
2- 3D measurements of irregularly shaped stems
3- 3D stem measurements at the plot level
4- Making tropical forest plots comparable
5- DiscussionNuméro de notice : 24037 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Thèse étrangère Note de thèse : Thèse de Doctorat : Sciences Agronomiques et Ingénierie Biologique : Liège : 2022 DOI : sans En ligne : https://hdl.handle.net/2268/293900 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=101855 Adaptation d'un algorithme SLAM pour la vision panoramique multi-expositions dans des scènes à haute gamme dynamique / Eva Goichon (2022)
Titre : Adaptation d'un algorithme SLAM pour la vision panoramique multi-expositions dans des scènes à haute gamme dynamique Type de document : Mémoire Auteurs : Eva Goichon, Auteur Editeur : Strasbourg : Institut National des Sciences Appliquées INSA Strasbourg Année de publication : 2022 Importance : 52 p. Format : 21 x 30 cm Note générale : bibliographie
Mémoire de soutenance de Diplôme d’Ingénieur INSALangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] cartographie et localisation simultanées
[Termes IGN] image panoramique
[Termes IGN] vision monoculaire
[Termes IGN] vision par ordinateur
[Termes IGN] vision stéréoscopiqueIndex. décimale : INSAS Mémoires d'ingénieur de l'INSA Strasbourg - Topographie, ex ENSAIS Résumé : (auteur) La Localisation et Cartographie Simultanées basée vision (SLAM) en robotique est bien établie mais trouve encore ses limites en environnement à grande gamme dynamique où les images acquises souffrent de sur- et sous-expositions. Ce travail s’appuie sur l’utilisation de caméras originales capables d’acquérir plusieurs expositions différentes simultanément en une image panoramique multiple pour limiter les saturations. Il en adapte les images et le modèle de projection en vue d’exploiter ces caméras dans le SLAM multi-caméra MCPTAM, initialement conçu pour des données différentes. Ce travail a permis de mettre en lumière les difficultés de MCPTAM dans les virages mais donne de meilleurs résultats avec des expositions multiples. Note de contenu : 1- Introduction
2- State-of-the-art
3- Description of methods used
4- Results
ConclusionNuméro de notice : 24092 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Mémoire ingénieur INSAS Organisme de stage : JRL (AIST-CNRS) / IRISA Rennes En ligne : http://eprints2.insa-strasbourg.fr/4672/ Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102562 Analysis of pedestrian movements and gestures using an on-board camera to predict their intentions / Joseph Gesnouin (2022)
Titre : Analysis of pedestrian movements and gestures using an on-board camera to predict their intentions Titre original : Analyse des mouvements et gestes des piétons via caméra embarquée pour la prédiction de leurs intentions Type de document : Thèse/HDR Auteurs : Joseph Gesnouin, Auteur ; Fabien Moutarde, Directeur de thèse Editeur : Paris : Université Paris Sciences et Lettres Année de publication : 2022 Importance : 171 p. Format : 21 x 30 cm Note générale : bibliographie
Thèse de doctorat de l'Université Paris Sciences et Lettres, Préparée à MINES ParisTech, Spécialité
Informatique temps réel, robotique et automatiqueLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Intelligence artificielle
[Termes IGN] apprentissage profond
[Termes IGN] attention (apprentissage automatique)
[Termes IGN] classification par réseau neuronal convolutif
[Termes IGN] classification par réseau neuronal récurrent
[Termes IGN] estimation de pose
[Termes IGN] image RVB
[Termes IGN] instrument embarqué
[Termes IGN] navigation autonome
[Termes IGN] piéton
[Termes IGN] reconnaissance de gestes
[Termes IGN] réseau neuronal de graphes
[Termes IGN] squelettisation
[Termes IGN] trajectoire (véhicule non spatial)
[Termes IGN] vision par ordinateurIndex. décimale : THESE Thèses et HDR Résumé : (auteur) The autonomous vehicle (AV) is a major challenge for the mobility of tomorrow. Progress is being made every day to achieve it; however, many problems remain to be solved to achieve a safe outcome for the most vulnerable road users (VRUs). One of the major challenge faced by AVs is the ability to efficiently drive in urban environments. Such a task requires interactions between autonomous vehicles and VRUs to resolve traffic ambiguities. In order to interact with VRUs, AVs must be able to understand their intentions and predict their incoming actions. In this dissertation, our work revolves around machine learning technology as a way to understand and predict human behaviour from visual signals and more specifically pose kinematics. Our goal is to propose an assistance system to the AV that is lightweight, scene-agnostic that could be easily implemented in any embedded devices with real-time constraints. Firstly, in the gesture and action recognition domain, we study and introduce different representations for pose kinematics, based on deep learning models as a way to efficiently leverage their spatial and temporal components while staying in an euclidean grid-space. Secondly, in the autonomous driving domain, we show that it is possible to link the posture, the walking attitude and the future behaviours of the protagonists of a scene without using the contextual information of the scene (zebra crossing, traffic light...). This allowed us to divide by a factor of 20 the inference speed of existing approaches for pedestrian intention prediction while keeping the same prediction robustness. Finally, we assess the generalization capabilities of pedestrian crossing predictors and show that the classical train-test sets evaluation for pedestrian crossing prediction, i.e., models being trained and tested on the same dataset, is not sufficient to efficiently compare nor conclude anything about their applicability in a real-world scenario. To make the research field more sustainable and representative of the real advances to come. We propose new protocols and metrics based on uncertainty estimates under domain-shift in order to reach the end-goal of pedestrian crossing behavior predictors: vehicle implementation. Note de contenu : 1- Introduction
2- Human activity recognition with pose-driven deep learning models
3- From action recognition to pedestrian discrete intention prediction
4- Assessing the generalization of pedestrian crossing predictors
5- ConclusionNuméro de notice : 24066 Affiliation des auteurs : non IGN Thématique : INFORMATIQUE Nature : Thèse française Note de thèse : Thèse de Doctorat : Informatique temps réel, robotique et automatique : Paris Sciences et Lettres : 2022 DOI : sans En ligne : https://tel.hal.science/tel-03813520 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102091 Deep learning based 2D and 3D object detection and tracking on monocular video in the context of autonomous vehicles / Zhujun Xu (2022)
Titre : Deep learning based 2D and 3D object detection and tracking on monocular video in the context of autonomous vehicles Type de document : Thèse/HDR Auteurs : Zhujun Xu, Auteur ; Eric Chaumette, Directeur de thèse ; Damien Vivet, Directeur de thèse Editeur : Toulouse : Université de Toulouse Année de publication : 2022 Importance : 136 p. Format : 21 x 30 cm Note générale : bibliographie
Thèse en vue de l'obtention du Doctorat de l'Université de Toulouse, spécialité Informatique et TélécommunicationsLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] apprentissage semi-dirigé
[Termes IGN] architecture de réseau
[Termes IGN] détection d'objet
[Termes IGN] échantillonnage de données
[Termes IGN] objet 3D
[Termes IGN] segmentation d'image
[Termes IGN] véhicule automobile
[Termes IGN] vidéo
[Termes IGN] vision par ordinateurIndex. décimale : THESE Thèses et HDR Résumé : (auteur) The objective of this thesis is to develop deep learning based 2D and 3D object detection and tracking methods on monocular video and apply them to the context of autonomous vehicles. Actually, when directly using still image detectors to process a video stream, the accuracy suffers from sampled image quality problems. Moreover, generating 3D annotations is time-consuming and expensive due to the data fusion and large numbers of frames. We therefore take advantage of the temporal information in videos such as the object consistency, to improve the performance. The methods should not introduce too much extra computational burden, since the autonomous vehicle demands a real-time performance.Multiple methods can be involved in different steps, for example, data preparation, network architecture and post-processing. First, we propose a post-processing method called heatmap propagation based on a one-stage detector CenterNet for video object detection. Our method propagates the previous reliable long-term detection in the form of heatmap to the upcoming frame. Then, to distinguish different objects of the same class, we propose a frame-to-frame network architecture for video instance segmentation by using the instance sequence queries. The tracking of instances is achieved without extra post-processing for data association. Finally, we propose a semi-supervised learning method to generate 3D annotations for 2D video object tracking dataset. This helps to enrich the training process for 3D object detection. Each of the three methods can be individually applied to leverage image detectors to video applications. We also propose two complete network structures to solve 2D and 3D object detection and tracking on monocular video. Note de contenu : 1- Introduction
2- Video object detection avec la heatmap propagation (propagation de carte de chaleur)
3- Video instance segmentation with instance sequence queries
4- Semi-supervised learning of monocular 3D object detection with 2D video tracking annotations
5- Conclusions and perspectivesNuméro de notice : 24072 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Thèse française Note de thèse : Thèse de Doctorat : Informatique et Télécommunications : Toulouse : 2022 DOI : sans En ligne : https://www.theses.fr/2022ESAE0019 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102136
Titre : Domain adaptation for urban scene segmentation Type de document : Thèse/HDR Auteurs : Antoine Saporta, Auteur ; Matthieu Cord, Directeur de thèse Editeur : Paris : Sorbonne Université Année de publication : 2022 Importance : 147 p. Format : 21 x 30 cm Note générale : bibliographie
Thèse de Doctorat de Sorbonne Université, spécialité InformatiqueLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] apprentissage profond
[Termes IGN] classification non dirigée
[Termes IGN] entropie
[Termes IGN] Mapillary
[Termes IGN] navigation autonome
[Termes IGN] réseau antagoniste génératif
[Termes IGN] scène urbaine
[Termes IGN] segmentation sémantique
[Termes IGN] vision par ordinateurIndex. décimale : THESE Thèses et HDR Résumé : (auteur) This thesis tackles some of the scientific locks of perception systems based on neural networks for autonomous vehicles. This dissertation discusses domain adaptation, a class of tools aiming at minimizing the need for labeled data. Domain adaptation allows generalization to so-called target data that share structures with the labeled so-called source data allowing supervision but nevertheless following a different statistical distribution. First, we study the introduction of privileged information in the source data, for instance, depth labels. The proposed strategy, BerMuDA, bases its domain adaptation on a multimodal representation obtained by bilinear fusion, modeling complex interactions between segmentation and depth. Next, we examine self-supervised learning strategies in domain adaptation, relying on selecting predictions on the unlabeled target data, serving as pseudo-labels. We propose two new selection criteria: first, an entropic criterion with ESL; then, with ConDA, using an estimate of the true class probability. Finally, the extension of adaptation scenarios to several target domains as well as in a continual learning framework is proposed. Two approaches are presented to extend traditional adversarial methods to multi-target domain adaptation: Multi-Dis. and MTKT. In a continual learning setting for which the target domains are discovered sequentially and without rehearsal, the proposed CTKT approach adapts MTKT to this new problem to tackle catastrophic forgetting. Note de contenu : 1- Introduction
2- Unsupervised domain adaptation
3- Leveraging priviledge information for unsupervised domain adaptation
4- Estimating and exploiting confident pseudo-labels for self-training
5- Adaptation to multiple domains
6- ConclusionNuméro de notice : 24079 Affiliation des auteurs : non IGN Thématique : IMAGERIE/INFORMATIQUE Nature : Thèse française Note de thèse : Thèse de Doctorat : Informatique : Sorbonne Université : 2022 Organisme de stage : Institut des Systèmes Intelligents et de Robotique DOI : sans En ligne : https://theses.hal.science/tel-03886201 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=102213 PermalinkExploring data fusion for multi-object detection for intelligent transportation systems using deep learning / Amira Mimouna (2022)PermalinkPermalinkOptimization of deep neural networks: A functional perspective with applications in image classification / Simon Roburin (2022)PermalinkPermalinkPermalinkScaling up and evaluating surface reconstruction from point clouds of open scenes / Yanis Marchand (2022)PermalinkPermalinkPermalinkPose estimation and 3D reconstruction of vehicles from stereo-images using a subcategory-aware shape prior / Maximilian Alexander Coenen in ISPRS Journal of photogrammetry and remote sensing, Vol 181 (November 2021)Permalink