Grégoire Maillet's research while affiliated with Institut national de l’information géographique et forestière and other places

... Pour les jeux de données synthétiques, deux bibliothèques spectrales sont considérées, chacune d'elle étant constituée de plusieurs signatures spectrales présentant le phénomène de variabilité intra-classe (plusieurs signatures spectrales sont disponibles pour un seul matériau). La première bibliothèque est construite via une campagne de terrain réalisée sur la ville de Toulouse (voir figure 4.1 [126][127][128]. Cette opération est réalisée à l'aide de la technique nommée Correction Code for Hyperspectral Images of remote sensing Sensors (COCHISE) [129]. ...

... It also demonstrates that an additional effort in data collection is worthwhile as the number of outputs to assess sustainability increases with the amount of inputs. In addition, with the development of new dendrometric parameter acquisition tools, such as drones, satellites and LIDAR [48,49], the possibility of obtaining accurate data over large areas will increase. The combination of these stand data with digital elevation models, soils maps, and regional climate forecast [37] offer a promising avenue for landscape simulation tools. ...

... The chosen urban dataset represents the city center of Toulouse (France), in spectral radiance. It has been acquired at a 1.6 m spatial resolution in the reflective range by the HySpex instrument from the 2012 UMBRA (ONERA-IGN) airborne campaign (Adeline et al., 2013), in the context of the HYPXIM/HYPEX-2/BIODIVERSITY hyperspectral mission (Briottet et al., 2017). This complete dataset (1417 × 1417 pixels) contains 408 spectral bands covering the reflective domain ([0.4 − 2.5 µm]). ...

... That approach is especially useful when it is possible to define geometric constraints that may be semiautomatically tested in the context of cadastral plans [Flamanc et al., 2003]. Specific urban survey experiences show that satellite images can be used to generate 3D models of dense construction complex [Maillet and Flamanc, 2004]. ...

... Hierarchical methods combine the local and global methods and take the advantages of both types of methods, such as the self-adaptive triangulation constrained image matching (SATM) (Wu et al., 2011(Wu et al., , 2012. However, most of the existing image matching methods are often not well suited to derive the 3D data in urban areas where usually tall buildings are presented, and much human interaction are still required when processing images in urban areas (Habib et al., 2010;Helpke, 1995;Paparoditis et al., 2001;Wu et al., 2011). This is mainly due to the image matching difficulties. ...

... Ainsi, au-delà du choix d'une méthode de relevé répondant aux contraintes inhérentes aux EES, l'expert doit pouvoir apprécier la qualité et l'adéquation de données tridimensionnelles et cela quels que soient leurs modes d'acquisition.). Les recherches actuelles en photogrammétrie digitale concernent notamment l'extraction automatique d'objets (bâtiments, routes, etc.) (Kim et al., 2001 ; Fuchs et al., 2002). Citons également l'utilisation d'autres vecteurs spatiaux pour l'obtention de données sur des zones plus restreintes, notamment l'emploi d'hélicoptères (). ...

... First of all, the stereo images need refinement and re-sampling works to make them suitable for 3D mapping purposes. The work of Cantou et al. represents an excellent example for the archaeological domain, since the image treatment depends on the material of the sites and the procedures cannot be extended to other domains [23]. Moreover, in some cases and depending on the purpose of the mapping, the resolution for the representation scale revealed insufficient [24], requiring different approaches or the integration with ground survey data. ...

... Fanfani & Colombo, 2019 discuss the detection of structural changes by exploiting vision-based 3D models of a time-changing environment to detect changes while (Nguyen et al., 2017) proposed an approach to compare CityGML models on both geometric and semantic levels by employing a graph database. Kaartinen et al., 2005, in their research, have compared accuracies of photogrammetric and laser scanning methods for building extraction and partly compared the methods used for the same. All of the research cited above and more focus on one aspect of modelling, a single type of representation or the method of generation. ...

... Much work has focussed on assessing the quality of the geometric aspects of models. This includes assessing the deviation of points in the model from the corresponding points in the verification data for height (Avrahami et al., 2005; Hu et al., 2005; Kaartinen et al., 2005; Taillandier, 2005), for horizontal distance (Avrahami et al., 2005; Haithcoat et al., 2001) and in three dimensional space (Alberto Guarnieri and Remondino, 2004; Hsieh, 1996a; Hsieh, 1996b; Jamet et al., 1995). Also, relative distances between points (or lengths of edges) are measured (Haithcoat et al., 2001; Hu et al., 2005; Kaartinen et al., 2005; Tucci et al., 2001; Ragia and Winter, 1998) (and Bell et al., 2003 for dental data). ...

... When architectonic typologies offer a limited universe of geometric hypotheses (e.g., roofs shapes), satellite images can also be used to fit objects into a limited list of choices [Flamanc and Maillet, 2008]. ...