期刊名称：Archiwum Fotogrametrii, Kartografii i Teledetekcji
印刷版ISSN：2083-2214
出版年度：2008
卷号：18
语种：English
出版社：Main Board of Association of Polish Surveyors
摘要：As part of this study,the authors have created a set and provided the methodology for acquiring hyperspectral data in terrestrial conditions. The set is based on a monochromatic digital camera with an optoelectronic tuneable filter. In the era of developing reconnaissance techniques,there is a need for conducting such surveys in real time. The systems used to capture imagery in real time (or close to real time) should assure a shortest possible time of hyperspectral image acquisition as well as its digital processing. For the European theatre of operations vegetation is usually the typical natural background. Therefore studies were carried out on the use of vegetation indices (e.g. NDVI) in object recognition from a natural background. The article presents research regarding the process of selecting appropriate hyperspectral scenes for a given vegetation index. The analyses were carried out based on hyperspectral images of real and artificial plants. The imagery was acquired at the Military University of Technology in Warsaw,with the use of the hyperspectral system created. Based on the results described in this paper,it is safe to say that it is possible to acquire and process images which can be used in the recognition of artificial objects against a natural background in almost real time,with the use of a hyperspectral technique. There is a need to continue research into the selection of a most appropriate band in which the difference between the object sought and its background are maximised. The use of an electrooptical tuneable VIS filter has limited the analysed range to 400-720nm. Using a NIR tuneable filter instead of the VIS filter would broaden the analysed range to 650-1000nm.
其他摘要：WyróĪnienie i identyfikacja obiektów wojskowych maskowanych i nie maskowanych jest gáównym celem wojskowego rozpoznania obrazowego. Rozpoznanie obiektów moĪe odbywaü siĊ przy wykorzystaniu technik hiperspektralnych,które umoĪliwiają tworzenie stycznych,ni
关键词：reconnaissance;hyperspectral technique;vegetation index
