<?xml version="1.0" encoding="UTF-8"?><xml><records><record><source-app name="Biblio" version="7.x">Drupal-Biblio</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Chirici, G.</style></author><author><style face="normal" font="default" size="100%">Scotti, R.</style></author><author><style face="normal" font="default" size="100%">Montaghi, a</style></author><author><style face="normal" font="default" size="100%">Barbati, a</style></author><author><style face="normal" font="default" size="100%">Cartisano, R.</style></author><author><style face="normal" font="default" size="100%">Lopez, G.</style></author><author><style face="normal" font="default" size="100%">Marchetti, M.</style></author><author><style face="normal" font="default" size="100%">McRoberts, R. E. E.</style></author><author><style face="normal" font="default" size="100%">Olsson, H.</style></author><author><style face="normal" font="default" size="100%">Corona, P.</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Stochastic gradient boosting classification trees for forest fuel types mapping through airborne laser scanning and IRS LISS-III imagery</style></title><secondary-title><style face="normal" font="default" size="100%">International Journal of Applied Earth Observation and Geoinformation</style></secondary-title></titles><keywords><keyword><style  face="normal" font="default" size="100%">airborne laser scanning</style></keyword><keyword><style  face="normal" font="default" size="100%">Classification and regression trees</style></keyword><keyword><style  face="normal" font="default" size="100%">Forest fires</style></keyword><keyword><style  face="normal" font="default" size="100%">forest fuel type mapping</style></keyword><keyword><style  face="normal" font="default" size="100%">irs liss-iii imagery</style></keyword><keyword><style  face="normal" font="default" size="100%">Mediterranean forests</style></keyword></keywords><dates><year><style  face="normal" font="default" size="100%">2013</style></year><pub-dates><date><style  face="normal" font="default" size="100%">2013///</style></date></pub-dates></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://linkinghub.elsevier.com/retrieve/pii/S0303243413000494</style></url></web-urls></urls><volume><style face="normal" font="default" size="100%">25</style></volume><pages><style face="normal" font="default" size="100%">87 - 97</style></pages><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">This paper presents an application of Airborne Laser Scanning (ALS) data in conjunction with an IRS LISS-III image for mapping forest fuel types. For two study areas of 165 km2 and 487 km2 in Sicily (Italy), 16,761 plots of size 30-m × 30-m were distributed using a tessellation-based stratified sampling scheme. ALS metrics and spectral signatures from IRS extracted for each plot were used as predictors to classify forest fuel types observed and identified by photointerpretation and fieldwork. Following use of tra- ditional parametric methods that produced unsatisfactory results, three non-parametric classification approaches were tested: (i) classification and regression tree (CART), (ii) the CART bagging method called Random Forests, and (iii) the CART bagging/boosting stochastic gradient boosting (SGB) approach. This contribution summarizes previous experiences using ALS data for estimating forest variables useful for fire management in general and for fuel type mapping, in particular. It summarizes characteristics of classification and regression trees, presents the pre-processing operation, the classification algorithms, and the achieved results. The results demonstrated superiority of the SGB method with overall accuracy of 84%. The most relevant ALS metric was canopy cover, defined as the percent of non-ground returns. Other relevant metrics included the spectral information from IRS and several other ALS metrics such as percentiles of the height distribution, the mean height of all returns, and the number of returns.</style></abstract><notes><style face="normal" font="default" size="100%">The following values have no corresponding Zotero field:&lt;br/&gt;publisher: Elsevier B.V.</style></notes></record><record><source-app name="Biblio" version="7.x">Drupal-Biblio</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Morsdorf, Felix</style></author><author><style face="normal" font="default" size="100%">Mårell, Anders</style></author><author><style face="normal" font="default" size="100%">Koetz, Benjamin</style></author><author><style face="normal" font="default" size="100%">Cassagne, Nathalie</style></author><author><style face="normal" font="default" size="100%">Pimont, Francois</style></author><author><style face="normal" font="default" size="100%">Rigolot, Eric</style></author><author><style face="normal" font="default" size="100%">Allgöwer, Britta</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Discrimination of vegetation strata in a multi-layered Mediterranean forest ecosystem using height and intensity information derived from airborne laser scanning</style></title><secondary-title><style face="normal" font="default" size="100%">Remote Sensing of Environment</style></secondary-title></titles><keywords><keyword><style  face="normal" font="default" size="100%">airborne laser scanning</style></keyword><keyword><style  face="normal" font="default" size="100%">Canopy proﬁle</style></keyword><keyword><style  face="normal" font="default" size="100%">cluster analysis</style></keyword><keyword><style  face="normal" font="default" size="100%">Gaussian mixture models</style></keyword><keyword><style  face="normal" font="default" size="100%">LiDAR</style></keyword><keyword><style  face="normal" font="default" size="100%">Multi-layered ecosystems</style></keyword><keyword><style  face="normal" font="default" size="100%">Shrubland</style></keyword><keyword><style  face="normal" font="default" size="100%">Supervised classiﬁcation</style></keyword><keyword><style  face="normal" font="default" size="100%">Vertical stratiﬁcation</style></keyword><keyword><style  face="normal" font="default" size="100%">Wildland ﬁres</style></keyword></keywords><dates><year><style  face="normal" font="default" size="100%">2010</style></year><pub-dates><date><style  face="normal" font="default" size="100%">2010///</style></date></pub-dates></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://linkinghub.elsevier.com/retrieve/pii/S0034425710000568</style></url></web-urls></urls><volume><style face="normal" font="default" size="100%">114</style></volume><pages><style face="normal" font="default" size="100%">1403 - 1415</style></pages><language><style face="normal" font="default" size="100%">eng</style></language><abstract><style face="normal" font="default" size="100%">Height and intensity information derived from Airborne Laser Scanning (ALS) was used to obtain a quantitative vertical stratiﬁcation of vegetation in a multi-layered Mediterranean ecosystem. A new methodology for the separation of different vegetation strata was implemented using supervised classiﬁcation of a twodimensional feature space spanned by ALS return height (terrain corrected) and intensity. The classiﬁcation was carried out using Gaussian mixture models tuned on a control plot. The approach was validated using extensive ﬁeld measurements from treated plots, ranging from single vegetation strata to a more complex multi-layered ecosystem. Plot-level canopy proﬁles derived from ALS and from a geometric reconstruction based on ﬁeld measurements were in very good agreement, with correlation coefﬁcients ranging from 0.73 (for complex, 3-layered) to 0.96 (simple, single-layered). In addition, it was possible to derive plot-level information on layer height, vertical extent and coverage with absolute accuracies of some decimetres (simple plots) to a meter (complex plots) for both height and vertical extent and about 10 to 15% for layer coverage. The approach was then used to derive maps of the layer height, vertical extent and percentage of ground cover for a larger area, and classiﬁcation accuracy was evaluated on a per-pixel basis. The method performed best for single-layered plots or dominant layers on multi-layered plots, obtaining an overall accuracy of 80 to 90%. For subdominant layers in the more complex plots, accuracies obtained were as low as 48%. Our results demonstrate the possibility of deriving qualitative (presence and absence of speciﬁc vegetation layers) and quantitative, physical data (height, vertical extent and ground cover) describing the vertical structure of complex multi-layered forest ecosystems using ALS-based height and intensity information</style></abstract><issue><style face="normal" font="default" size="100%">7</style></issue><notes><style face="normal" font="default" size="100%">The following values have no corresponding Zotero field:&lt;br/&gt;publisher: Elsevier Inc.</style></notes></record></records></xml>