标题：Development of Filtering, Segmentation and Modelling Modules for LIDAR and Multispectral Data as a Fundament of an Automatic Forest Inventory System
期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2004
卷号：XXXVI-8/W2
出版社：Copernicus Publications
摘要：For a multitude of planning tasks, e.g. planning of power-lines, urban- and landscape planning or for forestry time-consuming on site inventories have to be done and thereby large costs arise. Therefore it is desirable to automate the inventory as far as possible. To make this aim realistic software has been developed, which provide a fundament for such types of automatic systems. Three modules have been implemented and tested. First the DTM/DSM filtering – This task is mandatory to be able to estimate the height of all interesting objects. An algorithm has been developed based on an active contour model. It is similar to a method created by the Swedish Defence Agency. Secondly the single tree delineation is implemented as a process chain consisting of three main parts. Based on the DTM/DSM data and a pouring algorithm approximate tree crown boarder lines are calculated. Secondly geometric knowledge about form parameters and mutual geometric relations between tree regions are introduced to enhance the delineation. Finally a "ray-algorithm" is used to further improve the delineation result. The third module, the tree species classification consist on a linear stepwise discriminant analysis algorithm using both laser and multi-spectral data and finally a form fitting of extended superquadrics based either on raster or original laser data is tested as tool for the tree classification and modelling
关键词：Inventory; Forestry; Land Cover; Lidar; DTM/DSM; Multispectral; Superquadric; Fitting
