Hyperspectral remote sensing data with waveband width of nm level has tens or even several hundreds channels and contains abundant spectral information. Different channels have their own properties and show the spectral characteristics of various objects. Optimum bands selection and feature extraction from the varieties of channels are very important for the effective analysis and information extraction of hyperspectral data. This paper, taking Shunyi region of Beijing as a study area, comprehensively analyzed the spectral feature of hyperspectral data. On the basis of analyzing the information content of bands, correlation among different channels, band separability and spectral absorption characteristics of objects, a fundamental method of optimum band selection and feature extraction from hyperspectral remote sensing data were proposed.Three factors, the information amount of bands, correlation between bands and separability of objects in bands, are considered in selecting bands. The major steps of band selection are:1. Compute the correlation matrix of hyperspectral data, analyze the correlation between bands, and then according to the correlation partition the complete data set into three band groups. The bands in same group are highly correlated and the different groups are relatively independent.2. Considering that hyperspectral data has many channels which appear in groups, define the band index aspi=σiRw+Ra, in whichσiis standard variance of ith band,Rwis absolute value of average correlation coefficient between ith band and other bands in same group,Rais the sum of absolute value of correlation coefficient between ith band and all other bands in different groups.It is evident that with higherσiand lowerRwandRa, thePiis higher and the corresponding bandiis
better in whole. ThusPiis an important parameter in selecting band.3. Select several typical spectral classes as training samples, which are important objects to be classified in study area and have similar spectral feature, compute the separability of classes in different bands by Bhattacharyya distance.4. On the basis of band comprehensive evaluation by band index and separability, select optimum bands bearing abundant information and high separability.The method derived from hyperspectral data of Shunyi region was also applied to different types of hyperspectral data of other regions and similar conclusion was got. It shows that the proposed method in this
paper is of general significance.
“3S”techniques (Remote Sensing, Geographic Information System and Global Positioning System)has been used more and more in forest fire control in the world. MODIS is one detector of Terra' s (EOS-AM1), which has been lunched on December 18,1999 by NASA Goddard Space Flight Center of US. This document has studied the possibility of different bands to detect the burning information from MODIS data by analyzing the information of related bands' data. The method to detecting fire information from the data has also been studied. In this paper, we have effectively gotten the burning information from MODIS data by using the image enhancement techniques and multispatial resolution data. At the same time, the information of different spatial resolution data has been fused.
The combined navigation system has been among the main trends of the navigation system development because of its much better performance compared with the single navigation system. The achievements on the theory in the modern control field including the data processing manners of optimizing estimation theory has provided the basis for the realization of the combined navigation system. In this article, a method using Kalman combined filter as well as a classical combination method is introduced. And a comprehensive comparison is made between the all-combined filtering method and the distributed combined filtering method in which we keep our emphasis on the analysis and deduction of different combination manners of Kalman filter used in the software we produced.
It seems much difference on remote sensing images with different quality water. After analyzing
the relative spectral reflectance of different quality water on TM images, we found that the relative spec-
tral reflectance of band 1 to band 4 (R1, R2, R3 and R4) were sensitive to the different quality water. R2
will be greater than R1 (R2/R1>1) for high suspended sediment regions, in which, if the suspended sedi-
ment concentration is very high, R3 is greater than R2 (R3>R2), otherwise R3<R2. For contaminated
water, the more serious the polluted water is, the higher value R4/R3 is. According to the water monitor-
ing data from Environment Protect Bureau of Dongying City, the main pollutants of drainage system in
Yellow River Delta are organic substances. So R4/R3 may be used as an important index of water organic
pollution in the area. After numerous tries and comparisons, we found the average and standard deviation
of R4/R3 may be used as a semi-quality index for water pollution classification. The mouth of Xiao-qing
River was taken as an example to test the above rules and its water pollution distributions were got from
TM images in different years. The regions with high suspended sediment and serious polluted water can be
extracted from TM images using the method introduced above. And the method has some references for
water monitoring with large area.
This paper used the forest resources of Meicheng forest farm in Dongzhi county of Anhui province as example and built forest resources management information system based on Mapinfo4.0 and Viewgis2.0.The system can realize accurate query, modification, statistics analysis and provide visual results. The results showed that the structure of forest types and tree species are not rational. The ratio of timber forest type is more than it of economical forest, special forest etc.. The quality of forest stand is lower. Based on these analysis, the paper suggested that the ratio of ecological forest, landscape forest and broad-leaved forest should be increased respectively in Meicheng forest farm. In addition, the system can export different special subject maps such as maps of forest types distribution and trees species distribution ,and also forecast the dynamic state of forest resource ,provide the scientific base for enhancing forest management level.
Longkou City is located in the northwest of Jiaodong peninsula. It is one of the most prosperous areas in Shandong province. Because it is limited by the special natural condition, its per capita water resource occupancy volume is only 347 cubic meters, far below the average national level. And because the management means lags behind, the limited water resource can not be fully utilized. The issue of being short of water has already become the major obstacle of the sustainable development of the city' s national economy. Therefore, we devised the water resource geographic information system for Longkou City.This article introduces the aim, the constitution, the thought ways and the main functions of the system.Through a few years' application trial, the operation condition of the system has been proved good. It has made a contribution to local economic development.
Fuqing is one of the cities in the coastal zonal areas of Fujian province, showing the fastest eco-
nomic growth. The rapid economic development, especially in the 1990s, has resulted in the change of eco-
nomic structure of Fuqing from agriculture dominant to industry dominant. With this significant economic
shift, the area of urban/town land use has also increased. In essence, urban expansion in which non-urban
land use changes to urban land use is one aspect of urbanization. To reveal the actual change of this land
use type, remote sensing technology was employed, which extracted the urban/town land use information
from two Landsat TM images of 1991 and 1996, respectively. The study indicates that the urban/town
land use of Fuqing has grown by 16.17 km2during the five years, 1.2 times more than its original area in
1991. Obviously, the value of urban/town expansion is an important index, which is valuable for the as-
sessment of urbanization. The achieved spatial data of urban/town expansion were integrated with the so-
cio-economic data through a GIS. This finally reveals that the driving forces to the urban/town land use
expansion in Fuqing were the rapid industry development during the period.
Sustainable development (SD) has been accepted as a philosophical concept for long time. In this article, an integrated SD criteria system and an operable SD analysis method was set up on purpose of enable quantitative SD study of agricultural land. Firstly, data from RS, GIS and normal statistic department, which mainly focus on the spatial features of agricultural land, was integrated to construct a synthetical SD criteria system. Then, a series of plot-based spatial analysis and zonal statistics methods,which worked under ERDAS and ARC/INFO, was used to evaluate the spatial background feature of each land parcel. After that, SPSS was used to perform statistic analysis, such as dimension reduction and cluster analysis, on the large amount of dataset resulting from spatial analysis. Finally, the fast growing city of Guangdong Province, Dongguan was chosen as the sample area to evaluate its land-use change from 1988~1998. Some objective, quantitative and spatial-characteristic conclusion was calculated about the SD status of the agricultural land in Dongguan, These valuable results greatly enable our further understanding of the SD status of our living environment.
It is an engineering-oriented work to delimit the water resource protection area scientifically and
reasonably. According to the drinking water resource protection principle, with the water resource area of
Beixi Aqueduct and Bantou Reservoir in Xiamen as an example, the GIS technology of creating basin and
watershed automatically based on the foundation of Digital Elevation Model (DEM) from topography, the
remote sensing technology of acquiring natural ecological environmental information, and the GIS analytic
method of delimiting water resource protective scope making use of various data, including social and natu-
ral data, are studied in this paper. The focus is on the discussion of GIS methodology of obtaining bound-
ary of basin and watershed automatically based on the creation of DEM from topography. The required
natural ecological information for delimiting water resource protective area from remote-sensing are also
discussed and the puzzle in which is analyzed. However, applying remote-sensing and GIS technology to
delimit water protective is a kind of more scientific method to be performed than ever before and its result
is credible.
Monitor of dynamic change of land use is a strong means to master situation of resources and environment roundly and in time. Duty of monitoring of land use change was finished by manual work traditionally. So how to get large-area images accurately and promptly is very important for monitoring of dynamic change in regional scale. The main contents of this article are based on it. MODIS is a imaging spectroradiometer on a series of EOS-AM1 Satellites. In 1999, with the successful launch of EOS-AM1 satellite, the MODIS began to send remote sensing data in Dec, 1999. In this paper, we described the characters of MODIS data and compared it with other remote sensing data. Considering its advantages and disadvantages, we proposed the feasibility of using this data to monitor the dynamic changes in large areas and described the applications from three aspects in more details.
This paper proposes a new targets detection method based on MAX-Pyramid and affine motion modal for IR image sequence taken from moving platform. Firstly, we give a new pyramid structure to represent IR images-Max pyramid. The new structure, comparing with other pyramid structure existed such as Gaussian pyramid, is less computationally expensive and more suitable to represent infrared images. Then we use affine motion modal to describe the motion of platform, and the motion information of platform is removed after an affine image registration. Position and motion information of targets is detected in residual image. The result is also given in this paper.
Sustainable development strategy has become one of the most important theme in the world.The
theories,technologies and methods of remote sensing are widely applied in the development of the national
economy.They play an important role in the investigating,monitoring,analyzing,estimating and forecast-
ing of resources,environment and disasters.In this paper,land use change,spatial information resources
monitoring,soil erosion,ocean enviroment monitoring, national disasters monitoring, crop procduction es-
timating and enviromental pollution monitoring through remote sensing technology are discussed. Based on
above discussion,it is concluded that remote sensing technology enchances the feasibility of the realization
of global and regional sustainable development and will be one of the basic technology of sustainable devel-
opment study.
