We propose a new task, non localized scene graph verification, whose objective is to provide a justified expression of inconsistencies between the visual content of the image and its non-localized scene graph in order to diagnose errors or anticipate corrections. We introduce a sequential algorithm capable of detecting and proposing plausible corre...
The objective of this document is to introduce the principle of Autonomous PErceptual System (APES) as an object of study.The functionalities of artificial perception, in particular vision, have become both easier to design and more efficient through the use of a set of techniques and development environments grouped under the term "Deep Learning"....
For phenotyping operations, artificial vision devices operating in visible or hyperspectral color fields are used to perform geometric and colorimetric measurements. A robotic platform with a manipulator arm has been developed to detect diseases on sugar beet crops, during different growing stages. Image processing algorithms have been developed to...
Tout au long de l'année, des vidéos sont enregistrées en milieu forestier, sur différentes parcelles, pour détecter la présence d'animaux (cerf, sanglier, biche, pigeon,...) qui affectent les cultures dans l'environnement forestier. Ce travail de détection d'animaux est réalisé manuellement par des agents du site de Nogent, qui sont amenés à parcou...
Ce stage de fin d'études s'inscrit dans le cadre du projet Casdar Phenaufol qui concerne le développement d'un système robotisé embarquant un capteur de perception de type caméra couleur destiné
Le cadre de mon sujet s'inscrit dans le travail de suivi du dispositif expérimental OPTMix qui est un groupe de chercheurs IRSTEA basé à Nogent-sur-Vernisson. OPTMix vise à améliorer les connaissances sur le fonctionnement des forêts mélangées en région tempérée avec des applications directes à la gestion forestière en particulier dans le cadre des...
Stereoscopic reconstruction is important to automatic vision systems. As an intermediate step, estimating this reconstruction is not enough for good performance of the whole system, and its uncertainty must be characterized. Several methods propose uncertainty indexes based on specific data features, thus incomplete, while others are based on learn...
Precision agriculture vehicles need autonomous navigation in cultures to carry out their tasks. The detection of obstacles in crop row is crucial for safe navigation. Lidar devices or Time Of Flight cameras (TOF), allow to obtaingeometric data in natural environment. Fusion of those geometric information with a color camera data, improves the ident...
For automated maintenance operations for vines, a vision system, with one embedded camera on a mobile robot, was developed in order to detect and locate, in real time, vine trunks and to achieve automatic vehicle guidance close to the crop rows, to perform the task. The image processing algorithm developed consisted of a series of operations. Some ...
