| Based on the great advantages in the spectral resolution, high-resolution and hyperspectral remote sensing becomes milestone in the development of the modem remote sensing technologies. Hyperspectral image contains great consecutive band data of the ground surface once hyperspectral imager takes picture; it is not only the accumulation of the data volume but the addition of the spectral dimension, providing the abundant spectral data for the application of the remote sensing like observation of the ground surface, recognition of the ground object and surveillance of the geological and environmental change. In a word, this changes in natural the tradition of the application of the remote sensing.First, by methods of covariance matrix, optimal index and band index, the information content, the correlation of different channels and spectral character of ground objects are analyzed to extract band subsets with vast scale information content, smaller correlation, larger spectral difference using the complex ground surface hyperspectral image of nearby the Huangpujiang in Shanghai obtained by the airborne Pushbroom Hyperspectral Imager (PHI) to determine the optimal band combination of hyperspectral image. Then, based on the GVG instrument and transect sampling framework of a fast inventory method of crop planting structure, taking crop planting structure inventory for summer and autumn crop over China in 2002 for example, this paper introduces the method of combining sampling and remote sensing technologies that takes two-step sampling, and provides a substitute methodology to estimate crop acreage given that the real situation of the land use in China。Within the applications in two different fields, the remote sensing proves that it is of effectiveness, accuracy and economy. Last, we can integrate the hyperspectral technology with the existing remote sensing methods well and effectively promote the remote sensing technology of Geographic Information System。...
|
PDF Full Text Request