Based on the advantage of MODIS multi-spectrum data, this research explored a classificationmethod of feature selection and extraction, which combines the multi-spectrum data with multi-temporarydata in order to improve the classification accuracy. The classification accuracy was tested using 250 mMODIS data in Shandong province of China. The classification features were selected and extractedthrough the measures of the fractional cover, moisture, soil brightness, land surface temperature per day,and textures of various land cover types.The result indicates that it has higher classification accuracy using EVI (Enhanced Vegetation Index)as input than NDVI (Normalized Difference Vegetation Index), and NDWI (Normalized Difference WaterIndex) is superior to NDMI (Normalized Difference Moisture Index).The homogeneity of texture is thebest one for feature selection among the eight textures, and the optimal window size of texture is 11×11pixels. Whereas, NDSI (Normalized Difference Soil Index) almost has no effect for improving theclassification accuracy. For the contribution on improving the classification accuracy, EVI is the most, andthe following are homogeneity, NDWI and Tday(land surface temperature of day). The overall accuracyincreased about 10% through the method. The result shows that the feature selection and extraction canobviously improve classification accuracy, and the relatively high classification accuracy can also beacquired using the MODIS data sets without accessorial knowledge by this method.
In the paper, by means of statisticalFtest, reflectance spectra characteristics were compared andanalyzed synthetically among different varieties at different growth stages of summer maize, and statisticalmodels were established between red edge position (λred), red edge swing (Dλred), minimal swing230 (Dλmin),Dλred/Dλmin and Leaf total nitrogen (LTN), Chlorophyll content (Chl), Leaf area index(LAI). The results of study indicated: along with the growth, spectral discrepancy took on disciplinarychang, namely, the discrepancy gradually increased from jointing stage to tasseling stage, thereafter, thediscrepancy gradually decreased at anthesis and silking stage, but the discrepancy was bigger at anthesisthan that at anthesis, and yet the discrepancy surprisingly increased at milking stage.Dλred/Dλmin wasadded to simulate LTN more accurately thanDλmin during the growth, especially at silking stage,Dλredbefore anthesis,λred at jointing stage and bell stage. The result of utilizingDλred for estimating Chl wasencouraging before anthesis,λred at bell stage and tasseling stage. It was feasible to estimate LAI withDλred after tasseling stage or withλred andDλmin at tasseling stage, furthermore, usingDλred/Dλminwas more reliable than the others at anthesis and silking stage. Therefore, the proper red edge parameterswere so sensitive to corresponding agronomy index that they could be considered as responsive indicator ofthe agronomy indices of summer maize.
correlation analysis of the Landsat TM data, its derived data, meteorological data andtopographic data with the biomass of the tropical forest vegetation for the lithocarpus forest, the otherbroad leaf forest and the hard broad-leaf forest is explored here in Xishuangbanna, Yunnan province, P.R.of China. It includes four steps. Firstly, the biomass for each forest sample is calculated by using thefield inventory data of each sample. Secondly, GIS Database is established according the coordinate of eachforest sample. Secondly, Remote sensing image and its derived data, meteorological data, topographicaldata and the biomass of each sample are referenced to the same projection and coordination. Finally, thecorrelation between the Landsat TM and its derived data, meteorological data, topographical data and thebiomass is analyzed respectively for the lithocarpus forest, the other broad leaf forest and the hard broad-leaf forest. It is shown as follows: (1) The correlations of the biomass of lithocarpus forest and LandsatTM1,TM2,TM2,TM4,TM5,TM7,Bright,Green,Wet,VI3,DVI,PC1 and PVI are obvious at 0.05 level.(2) The correlation between the biomass of the other broad leaf forest and average rainfall per year isobvious at level 0.05. (3)The correlation of the biomass of the hard broad-leaf forest and average rainfallper year is obvious at level 0.05.Key words:
Forest fuel is a key factor of three factors of Forest Inflammation Cycle Theory. It' s the innerfactor of forest fire. It is important for Forest Fire Prevention, such as Forest Fire Danger RatingPrediction, Forest Fire Behavior Prediction (FBP), Forest Fire Extermination and Forest Fire Salvation.A method for classifying forest fuels by integrating MODIS data and GIS technique is tested in Heilongjiang province. In studying, five forest fuels have been classified after serials process, such as derivingthe Max Normalized Difference Vegetation Index (MNDVI ) over sixteen days' data by usingMultitemporal MODIS data, Analysising the MNDVI data by using Principal Components Analysis(PCA), Unsupervised classification method and Supervised classification method. It makes a strong basefor further studying National forest fuels classification method by using modernistic informationtechniques.
Land surface temperature can be used in the inspection of drought in the same vegetation index.In this paper, the principle of drought inspection using remote sensor was analyzed through thecharacteristic space of vegetation index and land surface temperature, land surface temperature wasreplaced with sensor temperature, the method of calculation drought index was discussed in detail. UsingLandsat7 ETM+sensor data in Luancheng County, the situation of drought was obtained and analyzed.
How to get high accuracy of classification by remote sensing image processing technology is oneof the difficulties in applications of wetland data. The paper discusses the methods to enhance the accuracyof classification for the wetland of Zhalong. First we can noise reductions enhance and tasseled cap enhanceto improve its accuracy. Then compare the classification results of original image with those of the othertwo enhance methods. The results are very similar, i.e., 83.91% of original image, 78.88% of noisereduction and 90.13% of tasseled cap respectively, checked by GPS sample points gained from fields, landcover data of 2000 and Kappa accuracy assessment. Each method has its own advantage in differentiatingthe classes, especially the accuracy in the wetland boundary after noise reduction enhance and the accuracyin the sub classes of wetland after tasseled cap enhance. But none of them is very satisfied. So based onknowledge and the technique of GIS, we can put the advantages of each methods together to gain anexcellent result. The new method is the kernel of this paper, using spatial algebra, which combines thetechnique of RS and GIS together, and it can greatly improve the accuracy and reflect the actual land typesbetter in classification of wetland. Its accuracy can get to 96% after checked by GPS sample points,landcover data of 2000 and visual interpretation. Then we can draw a conclusion that as far as wetlandclassification is concerted, the image enhance before classification can improve its accuracy in some parts,but using spatial algebra based on GIS technology can put the advantage together and get the best result.
Supermode is a completely original technique to improve the resolution of panchromatic images,there are some paper about this technique, but few regards the issue of Supermode image processing. wewill focus on this issue in this paper. Based on a brief introduction about the supermode adopted in SPOT5satellite, the sophisticated quincunx interpolation processing is considered, which can takes advantage ofthe regularity and structure inherent in interlaced data. Further more, some researches on denoising anddeconvolution have been proposed and the main factors affect the PSF of SPOT5 imaging system areanalyzed. Finally, we take some experiments on the SPOT5 supermode images processing and somesatisfactory results are achieved.
This paper introduces a real-time spaceborne SAR simulator and the hardware design based onthe PCI bus and CPLD chip. This simulator is mainly based on the parameters of the SAR and the satelliteorbit. The foundations of the simulation method are the point target model, the ellipse orbit model, andthe ellipse earth model. First, the simulation software can calculate both the phase-shifted data and theweighted data of the azimuth direction of antenna. These data can be synthesized and stored in one file.Second, these data are transferred to the time-controlled unit through PCI bus. Then, the LFM generatesthe linear frequency modulated signal, which is phase-shifted and weighted in the simulator, by means ofthe delay signal generated by CPLD timer. Finally, the system echo signal is formed. This signal can beused to verify the system characteristics, such as resolution, bandwidth, and the spectrum features etc..After digitalization, the data are stored in the storage device in the PC. Based on the classical R-Dalgorithm, the compressed data can be achieved. The compression results in this paper show that thismethod is feasible, and that the flexibility of the software and the hardware of the simulator hasadvantages in the system update and the distributed target simulator design.
The image feature fusion is a kind of data fusion that is based on spectral structure and texturefeature of the objects. This paper proposed feature fusion after enhancing image edge with wavelet’smulti-resolution analysis to largely improve image' s definition and resolving power. In the fusion process,instead of using the conventional HIS-Wavelet fusion, directly introduced RGB-Wavelet transformationwhich decomposed RGB bands of the multi-spectral image and high-resolution image by waveletseparately, then used low frequent parts of the R, G, B bands and high frequent parts of the high-resolution image to do the image fusion, finally had new R, G, B bands as the new fusion image. In theway, not only improved definition and resolution of the multi-spectral image, but also retained the colorfeature of multi-spectral image which was important for multi-spectral fusion. At last validated theconclusion with an experiment. This method was based on wavelet multi-resolution analysis and RGB colorspace, so it didn' t generate any compute error of color space transform. Compared the fusion results of RGB-wavelet and HIS-wavelet methods, it could find that not only precision of visual estimation but alsoguide lines of quantitative analysis such as definition, space resolution had greater improvement.
In this paper, data models of returned chirp signal for Spotlight SAR are inducted and described.In this model, returned signal consists of two parts——the returned sinusoids signal of dominant scattersand the additive noise and clutter. The sinusoids have a discrete spectrum, while the noise and clutter hasa continuous spectrum. Based on such model, the RELAX algorithm is introduced in the feature extractionof targets which can be divided into several dominant scatters: estimating these dominant scatters’discretespectrum parameters from the noise and clutter’s continuous spectrum. Among the estimated parameters,the frequencies correspond to the location of dominant scatters, and the amplitudes are proportional totheir RCSs. By this means, we can realize target feature extraction for this specific kind of targets.Finally, computer simulation results are presented to show the validities and advantages of RELAXalgorithm.
In the real-time processing system of Synthetic Aperture Radar, the preprocessor play a veryimportant role and the main component of the preprocessor is FIR filter. As for the requirements of theSynthetic Aperture Radar' s preprocessing filter, this paper proposes a method of FPGA based FIR filterdesign by linear FIR digital filter window function and accomplished a 16-tap 8-bit FIR filter by usingXCV100E FPGA devices. As for the key technical problem—the multipliers, we give the DistributedArithmetic which converts the multiplications into some look-up operations. This arithmetic uses the LUTresources in the FPGA fully and can improve the system' s speed and system' s reliability. The designmethod is very useful for the accomplishment of real-time processor of SAR and will benefit the space-borne Synthetic Aperture Radar real-time processing system. The results of simulation on the hardwarecircuit prove this method is efficient and can meet the design requirements very well.
Meteorological satellite is main detecting tool in distinguishing weather phenomena such as duststorm, considering both time resolution and space resolution. In order to exert the effect of themeteorological satellite mostly on distinguishing and forecasting dust storm, channel selection of detectingdust storm is the key point. By applying Mie theory to calculate scattering and extinction characteristics ofdust particles, we analyze dust channel selection of meteorological satellite according as the extinctioncharacteristics. According to calculation, visible channel is the main channel in detecting dust particles.Short-wave IR also has some sensitivity to dust observation. At the same time, the radiative temperatureof dust particles in shortwave infrared and longwave infrared is different mainly because emissivity is thefunction of wavelength. We consider to apply the difference of two infrared wave bands to distinguish duststorm.
Random phase error in the radar data will badly influence the pulse compress effect and blur theimage, due to the difficulty of removing this multiplicative phase noise. Document〔5,6〕threat the noise bycomputation of the high order statistics of received. However, this method requires large number ofsampling data to satisfy the stability of the statistics, which limits the application of this method.Document〔7〕brought out a new way to enhance Chirp signal from strong additive background randomnoise by rotating the short time Fourier transform of the signal, singular value decomposition and lineapproximation.This paper analyzes the influence of random phase noise from azimuth on SAR imaging first. Then onthe base of document〔7〕, a novel and effective way of rotation of Ambiguity Function and singular valuedecomposition to take place of matched filter is put forward, which is applicable to moving targetsdetecting and imaging since it is always presumed several strong reflecting point from the moving target.Simulation results are presented to confirm the theoretical results, showing that the method in this paper ismore effective than filters in time or frequency domain. However, a shortcoming is that the number ofecho signals will affect the filtering result by this method.
This paper presents research of several calibration methods on secondary planet, according toprinciple and characteristic of the Interference Super-spectral Imager, and requirement of calibration onsecondary planet. A method of Calibration based at first image plane makes use of two ends of narrow slitor a hole beside it to introduce the calibration light. There are two methods of converging calibration light,one is using convergent lens, the others self-focus lens. In method of whole system calibration, calibrationlight can be introduced by rotating the mirror 90 degree. Lambert reflector or integral sphere can be usedin this calibration. Place Nd-in glass in optical circuit can achieve both radiometric calibration and spectralcalibration. In actual, Analyzing the structure and characteristic, according to optic design and realizationpossibility of structure, a method of integral sphere used in calibration is chosen. By contrast, acalibration method of using integral sphere has compact structure and large energy, can be used to achievethe calibration of total system, aperture, and part viewing field.
We fully discuss algorithms of Kriging interpolation, isoline drawing and contour map fillingwhich are important in scientific data visualization and geographical information systems. Though thesetopics are classical in the theory of geographical systems and have undergone intense studies for manyyears, these algorithms are not easily found elsewhere in a so detailed and complete form, notably thecontour map filling is devised totally through our own much effort. To be more practical, we also presentthe popular DLL-and-XML-technology based architecture of a real rainfall analysis system which has beenused effectively in many cities of ZheJiang province. This greatly eased the hard task of rainfall isolinedrawing, formerly furnished mainly by hand. The sources of input rain data are vast, ranging fromrealtime recording at rain stations to the more advanced radar measurement, as long as these data werestored in databases or files. Both the algorithms and the framework of our software can be used withoutchange to other circumstance such as mountains' height contour and mineral distributions. The algorithmsintroduced here are robust and the overall functions of this software can rival those of the famous generalpurpose GIS software-Surfer by Goldensoft Corp. both in speed and appearance.
Oil spill has tremendous effect on marine environment. Therefore, it is significant for protectingmarine environment to get oil spill information timely. Presently, the methods which most countries adoptto detect oil spill are mainly direct detection or remote sensing methods. Among these methods, Synthetic Aperture Radar (SAR), one technique of remote sensing methods, has been becoming a hot research field nowadays. This paper concludes main characters of SAR images in oil spill detection. It introduces,compares and analyses the oil spill detection steps and realized techniques. There are four steps in oil spill detection process, including filtering, target detection, feature extraction and classification. Filtering canbe completed by several filtering techniques, such as Lee, advanced Lee and Frost. General speaking,there are four methods can realize target detection, which are single threshold, adaptive threshold,wavelet transform and max entropy method. About the third step, feature extraction, 12 features can beused to represent oil spill. Classification is a key and difficult step in oil spill detection, in whichalgorithms based on Bayes and neural network can be adopted. Finally, the paper points out thedevelopment direction of oil spill detection in SAR images.
