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：2002
卷号：XXXIV Part 4
出版社：Copernicus Publications
摘要：New high-resolution satellite images such as those from IKONOS provide potentially useful information for the identification of individual surface objects such as trees and buildings compared with previous satellite image sources (e.g. SP OT, Landsat). However, the detailed structure of buildings and trees is still ambiguous for the application of existing machine vision algorithms, mainly due to the resolution limit, currently about 1 metre, the existence of strong shadows due to the large oblique angles and the local time of overpass. Other data sets such as Lidar, aerial photographs, GIS information and pre-existing DEMs, can be used to complement these IKONOS image drawbacks. In particular, owing to the current paucity of IKONOS stereo-pairs, it may be possible to utilise Lidar data as an alternative source of 3D information, where available. This hypothesis has been tested with a 1m pan sharpened multispectral image of East London using a geometrically corrected IKONOS image and coincident 2m Lidar data using a new algorithm which identifies regions of interest (ROIs) for landscape objects (such as trees & buildings). This new algorithm was developed using a Bald Earth Digital Terrain Model (DTM) derived from Lidar and NDVI derived from IKONOS. As a result, excellent building outlines and individual tree extents were detected. However, results shown here applied to an IKONOS stereo- derived DEM indicate that the existing quality of such DEMs is not yet good enough for the detection of small objects.. A new processing system to refine IKONOS DEM quality is currently being developed and higher resolution satellite images (such as Quick-Bird) have recently appeared leading to the expectation that high resolution satellite based GIS construction may be realised in the near future
关键词：High resolution satellite image; DEM; DSM; Lidar; automated Building and tree detection
