One of the most important tasks in analyzing hyperspectral image data is the classification process. In general, in order to enhance the classification accuracy, a data preprocessing step is usually adopted to remove the noise in the data before classification. But for the time-sensitive applications, we hope that even the data contains noise the classifier can still appear to execute correctly from the user’s perspective, such as risk prevention and response. As the most popular classifier, Support Vector Machine (SVM) has been widely used for hyperspectral image classification and proved to be a very promising technique in supervised classification. In this paper, two experiments are performed to demonstrate that for the hyperspectral data with noise, if the noise of the data is within a certain range, SVM algorithm is still able to execute correctly from the user’s perspective.
Regular monitoring and assessment on radiometric performance of satellite sensors is necessary for the quantitative remote sensing application and development. HJ satellite was launched by China on 2008, which was put forward to achieve dynamic monitoring of environment and disasters, also need to monitor radiometric performance and provide stable and reliable calibration coefficients timely. In this study, Terra/MODIS data were used to calibrate HJ-1A CCD camera by cross-calibration technique in Dunhuang radiometric calibration site. Total thirteen HJ -1A CCD images were utilized, the 6s model was used to estimate the spectral matching factors. Finally, this study obtains long-term HJ-1A CCD calibration coefficients from 2009 to 2012. Results show that each band of HJ-1A CCD is varying degrees of degradation after HJ satellites launched 5 years later. This study is helpful to obtain high accuracy and reliable calibration coefficients and monitor radiometric performance of HJ-1A CCD.
With the deterioration of water pollution, monitoring of water environment is becoming more and more urgent.
However, there is no professional water environmental monitoring system in China. To overcome these problems, we
have developed a Surface water environmental monitoring System (WATERS for short) by VISUAL C++6.0 IDE.
WATERS is designed for the four kinds of remote sensing data of HJ-1 satellites, which are multi-spectral camera, ultraspectral
imager, infrared camera, and SAR. Besides, WATERS can also support other satellite remote sensing data. We
use some simulated HJ-1 satellites remote sensing data, as well as remote sensing data of similar satellite sensors, to test
the operation of WATERS. The operation results by these remote sensing data show that WATERS works well, and both
the efficiency and the precision of water quality monitoring are high.
Remote Sensing Image Simulation, which provides the image according to the characteristics of sensor in geometry, spectral and radiometry, is a very significant issue before the satellite's launching. This paper is mainly detailed in the geometric characteristics of multi-spectral sensors and the process of image geometric simulation of 1A microsatellite in Environment and Disaster Monitoring Microsatellite Constellation (HJ-1A). The relationship among the swath, the convergent angle, and the Ground Sampling Distance (GSD) is studied. The ideal photography model based on the photogrammetric theory is set up according to part of orbit parameters and reasonable assumptions. The relationship between the image coordinates and the geodetic coordinates and the simulating image algorithm is put forward. Finally, some conclusions are drawn up from the results of the study.
Quantitative remote sensing and GIS technologies are providing opportunities for Bioproduction and natural resources monitoring. Some gross primary production (GPP) and net primary production (NPP) models take ecological structure parameters (Land cover, LAI, and biomass) retrieved from remote sensing data as inputs to get information on ecosystem exchange on a global scale. Our recent research focuses on key problems in retrieving variable parameters of land surfaces from remote sensing observations, such as, the observing scale is different from the measuring scale of ground truth, the parameters which application requires may not be the same as the parameters of current physical models, the uncertainty of retrieved parameters by model inversion. In this paper, we present our new approaches on scaling effect modeling, physical model inversion by using prior knowledge, field experiments and setup of the spectrum knowledge base of typical objects, and the applications of the quantitative remote sensing research achievements in ecosystem assessment and monitoring, such as, in GPP and NPP of forest, energy exchange of crop field and grassland, water cycle, climate change. Some new modeling ideas and parameters retrieving results will be shown in this paper, as well as some remote sensing application samples.