Weather radar plays an important part in modern comprehensive meteorological observation system.Radar network can cover large spatial dimension,which is beneficial for the data fusion and elaborately observe different scale weather phenomenons.This paper summarizes the weather radar network technology and its application progress,including the bistatic radar network,Americar NEXRAD and TDWR network,European OPERA plan and china CINRAD network,finally the ultramodern CASA radar network.An exhaustive introduction about CASA network is presented,including the end-to-end system,DCAS mode,data product and specification comparison with NEXRAD.CASA can depict more detail change of the meso microscale weather,which provides helpful supplement to for large radar network.The mutual complementation and data fusion of the high-power,long-range S/C-band radar network and the low-power,short-range X-band radar network are the development tendency in the future.
Many methods have been proposed for building area extraction and building 3-D reconstruction from SAR images.However,the basic theory and process are diverse.To get better result,it is important for these methods to make full use of the information,including amplitude,polarization,and phase in SAR images.From this point of view,this paper reviews the current techniques and methods sequentially.The methods used for building area extraction are introduced by the order of used middle-resolution image,high-resolution image,time series images,fusion of SAR images and optical images.The methods used for building 3-D reconstruction are classified as radar photogrammetry and radar interferometry.Finally,some comments and promised ideas are given such as building height retrieval based on precise electromagnetic calculation and simulation,building area extraction based on accurate polarimetric decomposition techniques,and SAR image analysis based on new techniques from computer vision.
Intercalibration of the thermal infrared window channels 1(~10.9 μm)and2 (~11.9 μm) of the Stretched Visible and Infrared Spin Scan Radiometer on FengYun 2 (SVISSR/FY-2) satellites C,D and E with the hyper-spectral channels of the Atmospheric InfraRed Sounder on board Aqua (AIRS/Aqua) was developed using the High Spectral Convolution (HSC) method,which was improved in two aspects,sub-pixel spatial matching and automatic removal of the measurement pairs with coarse errors.According to the view geometries of SVISSR/FY-2 and AIRS/Aqua,with the study area longitude from 41.5°E to 149.5°E and the latitude from 10°S to 10°N was selected.To transfer the calibration coefficients from AIRS/Aqua channels to SVISSR/FY-2 channels,the SVISSR/FY-2 Normalized (NOM) 1B data and AIRS/Aqua 1B infrared product in December of 2006,September of 2007,December of 2007,September of 2009 and May of 2010 covering the study area were employed and aggregated with longitude and latitude spatial resolutions of 0.5°×0.5°.Measurement pairs are extracted from the aggregated SVISSR/FY-2 and AIRS/Aqua images with the following criterions:co-location,absolute solar zenith angle difference in the same day less than 2°(|ΔSZA|<2°),absolute viewing zenith angle differences less than 3°(|ΔVZA|<3°),and absolute relative azimuth angle differences less than 6°(|ΔRAA|<6°).The results show that the SVISSR/FY-2 measurements are linearly related to the convolved AIRS/Aqua measurements with large correlation coefficients.Conversely,AIRS/Aqua channels,the calibration biases exist in the two SVISSR/FY-2 thermal infrared channels,and the re-calibration coefficients are presented.
When the electromagnetic wave of the satellite radar altimeter get through the ionosphere in the atmosphere,the spread time will be postponed by the influence of the ion,thus the altitude data of the satellite radar altimeter measured could cause an error,as to it has to carry on correcting.This paper compared the main ionosphere correction models for the satellite radar altimeter,including Bent model,IRI95 model and IRI2007 model.The pros and cons of Bent model,IRI95 model and IRI2007 model are analyzed from three different angles which are layering,the mathematic model of each layer and the parameter established respectively.This study shows that IRI model performs better than Bent model,IRI2007 is better than IRI95,particularly in the medium degree of latitude region,the correct result of IRI2007 model is the best,based on the data of Jason-1 altimeter.The study provides the reference for the model selection of Chinese HY-2 altimeter.
This paper proposes a river detection method based on sub-block histogram and regional characteristics for the river extraction in high-resolution SAR images with complex background.Firstly,using wavelet transform reduces speckle and at the same time keeps the edge features of the image.Then utilizing statistics histogram initially fix the markers of river on the sub-block image.On that basis,it executes the marker-based watershed transformation to acquire the initial segmentation of the river.Finally it gets the detection results by the regional adjacency graph (RAG).Compared with the algorithm proposed by Guan-Zhao Dai,etc.which used the statistics from histogram for fast segmentation the method in this paper that can quickly determine the position of the river and extract the river body with higher extraction accuracy.This method also can be further used for bridge detection.The experiments show that the good usability method can effectively extract the outlines of river in high-resolution SAR images with complex backgrounds.
In order to effectively reduce the speckle noise of SAR intensity image which degrades the SAR image quality seriously,a new algorithm is proposed combing the improved Sigma filter with Gamma MAP filter.Firstly,high-return point targets are detected and preserved by using the threshold detection method.Secondly,sigma range is defined according to the distribution model of SAR intensity image and the MMSE estimator in a local window,the size of which can be adjusted adaptively through the local statistic property.Finally,the pixels within the sigma range are selected in a larger window and used to estimate the central pixel with the method of Gamma MAP filter.Both the spaceborne SAR and the airborne SAR images are filtered by the proposed algorithm compared with common filtering algorithms.The results show that the proposed algorithm excels in both the suppression of the speckle noise and the preservation of the texture details in edges,therefore the method can greatly enhance the ability of SAR image interpretation and target recognition.
This paper presents a new method for Landsat TM Image processing which can be used to quantitative simulating each band reflectance of TM image.On the basis of non-destruction of the original band spectral features,it improved single band information and visual sensitivity.The restructured Image quality were analyzed by image quality assessment indexes,and the results show that the image enhancement change and spectral information optimization degree of reconstructed Image have greatly improved.
Taking remote sensing image cognition and geographical understanding as the main analytic perspective,this paper firstly explored the characteristics of the image objects spectral,shape,textural and semantic by the multi-scale segmentation,to determine the corresponding relations between the characteristic information of the object and the ground.And then,this study rationally selected some classification features and built classification rules.Finally,multiple level classifications were carried out hierarchically to extract information of the interested objects in the study area.The result showed that the selected features in classification rules not only can effectively extract information,but also had the explicit geographical significance.Compared with the traditional pixel-based maximum likelihood classification methods,the classification accuracy of the new method was improved significantly.
In November 5th 2010,new generation meteorological FY-3B satellite successfully was launched,Microwave Radiation Imager (MWRI) can all day get radiation information from earth surface and atmosphere.Two scan periods were set during orbit test,1.7 s and 1.8 s separately.In this paper,image quality differences were analyzed in detail.On the study of image quality evaluation,statistical method was used to compare the dynamic range of image.By using power spectrum,space texture of image can be contrasted.Whats more,information entropy was used as a tool to evaluate the information magnitude.Finally,the image contrast and match situation among different channels were studied.The results show that 1.7 s is obviously better than 1.8 s in space texture,channel match and contrast,which means the image quality of 1.7 s is better than 1.8 s.The conclusion can be an important reference of MWRI instrument parameter design.
In view of such problems as the poor contrast,low resolution in remote sensing image,and to deal with the illumination distribute inharmonious and the “graying out” phenomenon existing in traditional Retinex remote image enhancement algorithms,a new remote sensing image enhancement algorithm based on multi-scale Retinex and histogram equalization is proposed in this paper.Firstly,the nonlinear global improvement of multi-scale Retinex algorithm is used to enhance V component in HSV color space.Then, the three components of RGB space will be synchronous enhanced processing with histogram equalization method. The experimental results show that images entropy increased by 29.5% compared with the multi-scale Retinex algorithms.It also shows that the algorithm both can restore the remote sensing image illumination and solve the “graying out” phenomenon.
Hyperspectral images segment algorithm is proposed through fractal feature image of original spectrum curve and high-frequency coefficients from spectrum curve wavelet decomposition.Fractal measurement of hyper spectrum curve is analyzed.A fractal dimension calculation algorithm with high-frequency coefficients is proposed to obtain multiple scale fractal feature images.A flow chart of hyperspectral image segment combined feature images with the wavelet coefficient of spectral response curve that is designed.Experiment results of provided hyperspectral images segment algorithm show that it can obtain the same image segment result but the higher efficiency is better than the direct fractal dimension feature image of spectral response curve.
As a mature image segmentation method,Multi-scale segmentation is used widely in remote sensing image information extraction,but the whole algorithm efficiency is low.In this paper,data paralleled segmentation of remote sensing image is achieved based on multi-core computer.The traditional image IO method can't satisfy the needs of parallel computing based on multi-core computer.A new image IO strategy is devised to eliminate this drawback;In addition,in the process of parallel segmentation,it universally exists a problem that the segmentation results can not be merged directly.The method of re-segmentation is used to solve the problem to the special areas.The experimental results show that parallel segmentation efficiency has improved significantly on images of various quantity and size.Besides,the segmentation algorithm has been provided with the ability of processing image of large data quantity and the applicability is greately enhanced .Image segmentation based on multi-core computer acquires remarkable effect.
As a source of new satellite data,Medium Resolution Spectral Imager (MERSI) on board the second Generation Polar-Orbiting Meteorological Satellite of China (FY-3) is not only featured with its finer surface-observing ability,but also its multi-channel and huge amount,which takes too much time and frequent I/O operation to correct by Commercial RS software.This paper proposes a flowchart to fast pre-process FY-3/MERSI data,the key steps of that are,according to the study area's shapefile and FY-3/MERSI self-positioning data,the corresponding region images of all 20 channels are georeferenced using geometric correction algorithm based on triangular network,followed by radiometric calibration,solar elevation angle revision,and so on.The series of procedures are implemented based on IDL program,and realize module interface for user interaction.Test results show that the algorithms in this module take less memory and improve the efficiency of FY-3/MERSI data processing greatly,for an area of about 450 000 km2,it takes only 450 s for all channels,significantly faster than using geographic lookup table (GLT) algorithm in ENVI software to achieve geometric correction,which takes over 30 min for only one channel.Besides,the quality and geographic accuracy of the corrected images are better.The development of this module greatly enhances FY-3/MERSI data processing efficiency,and it completes the prerequisite work for FY-3/MERSI application in thermal environment,snow and other various fields.
In order to improve the classification accuracy of satellite cloud imagery and identify cloud types in real time,a knowledge-based object-oriented classification algorithm was proposed.Firstly,500 cloud samples were cut from FY-3A VIRR data obtained in July and August 2011 after preprocessing,and 42% of them were randomly selected for training samples.Spectral and textural features were extracted from each of training samples followed by feature selection using ReliefF,and the back-propagation neural network was trained to build the classifier,and tested on the rest of 58% of cloud samples,then the knowledge base of cloud types was established.To interpret a whole satellite cloud image,the JSEG segmentation method was employed to derive objects,and the trained classifier in the knowledge base of cloud types was applied to the object-oriented classification procedure.Experimental results show that the proposed algorithm for satellite cloud image classification is effective,and the generated classification results are basically consistent with the corresponding cloud classification products.
Differential Interferometry SAR(DInSAR) technique has been widely used in surface subsidence survey.This paper introduces the use of differential InSAR technology to monitor surface subsidence in mining area.Both C band Envisat ASAR and L band ALOS PALSAR data were used to generate deformation interferograms.The characteristics of coherence as well as phase patterns on the C band and L band deformation interferograms were comparatively analyzed.The experimenta results indicate the feasibility of using multi-mode radar data for monitoring the mining area subsidence by considering the advantages and disadvantages of C band and L band in combination.Meanwhile,through the analysis of deformation interferograms,which can get the accurate location of ongoing mining area.
Environment and Disaster Reduction Satellite Constellation has the ability of wide coverage and rapid revisit,so that it can fulfill the demand of large scale disaster observation such as drought.Based on the weight overlaying method,regional drought risk assessment method comprehensively considers disastrous factors,disaster formative environment,hazard-bearing body and other factors,and achieves drought risk level distribution map which can be used in early-warning of drought.Spectral dimension-temperature drought index takes into account of four factors,including soil moisture,surface evaporation,vegetation greenness and plant moisture changes,so it can be applied in drought monitoring.For serious drought of Southwest China in early 2010,this paper carries out drought early warning and monitoring using Environment and Disaster Reduction Satellites data,and verifies risk assessment and monitoring results using local government reported disaster information from NDRCC,which shows that the results are in line with the actual disaster information with high accuracy.Risk assessment and monitoring results still need further improvements to meet the accuracy requirement of natural disaster relief emergency plan simulation and emulation.
Training sample size,auxiliary data and classification algorithm are the three main factors influencing on the land use and cover classification accuracy,through finding out the best combination of the three factors to improve the classification accuracy.Multispectral imagery and DEM from ETM satellite are applied to perform experiments under four sample sizes (100%,75%,50%,25%),along with calculated NDVI layer and mean texture measure.By contrasting the classification accuracy produced using three different classification algorithms:Classification and Regression Tree(CART),Support Vector Machine(SVM) and Maximum Likelihood Classification(MLC),to discuss the influence of training sample,auxiliary data and different classification algorithms to the classification accuracy.The results indicate that the SVM can obtain the best classification result when there is no effective auxiliary data,and the total accuracy can reach to 85%,when CART is used to classification ,as the additional NDVI and texture feature,the accuracy increased 2.82%,it is sensitive to auxiliary data ,which can obviously improve the classification accuracy.While in the condition of limited training sample obtain the effective auxiliary data,the CART can achieve the highest classification accuracy.
Relationship between landslide and topographical factors is one of the focuses of study the law of regional landslide;however,previous research mostly focused on its relationship with factors reflected terrain geometry form(Geometry Attribute Factor).Generally,there are two kinds of methods for describing topographical factors.Taking the slope position as an example,and it can be depicted by classifying slope position according to geometry attribute factors;and it can also be depicted by some kind of index calculated according to regional relative location information(Relative Position Factors).This paper first proposes a new method to depict slope position quantitatively based on Height Above the Nearest Drainage(HAND).HAND is defined as the elevation difference between a point and corresponding outlet through the point which drop the water into the river.And the quantitative slope position is the ratio between the points HAND value and maximum elevation difference within the sub basin.The another research is to examine the relationship between landslides and two different types of topographical factors named geometry attribute factors and relative position factors.The research area is located in Ou River in Wenzhou area and the landslides occurred during 1990-2001.The relationship is calculated by the weights of evidence method,which is a good candidate for calculating the landslides and some single topographical factor.The results show that higher correlation exits between relative position factors and landslide than geometry attribute factors which are common used in the risk evaluation of landslide,which indicate that the regional relative position information is more related to the landslide than the simple basic terrain geometry information,and most of the landslides occurred in relatively-low-HAND region in the watershed,which is suffering from water rushing with greater intensity and more frequency.
The technology in optical satellite remote sensing develops quickly,which is important to seek how to detect the ship target from large amounts of optical satellite remote sensing images.Aiming at the problem that traditional CFAR detection cannot detect the black polar target,a ship target detection method based on GLRT (Generalized Likelihood Ratio Test) is proposed.Like CFAR detection,this method uses a sliding window to detect target.Assuming the intensity of the background and target follow Gaussian distribution,detects the target through the judge that the distribution in the background window and target window is the same by the GLRT theory,which considers both the white and black polar target.Detection is operated in the image blocks and morphological operation is used for target cluster.The images used in Experiments come from SPOT5 and CBERS.Assume that the intensity of the background follows Gaussian distribution is validated.The result of detection in two images shows that this method can detect the black polar target and have fewer false alarms than CFAR detection.The ROC curve and the cost time that is 40% less than CFAR detection,and further show that this method is more superior.
Terrain is an important factor for affecting the distribution and vegetation changes.It is very important that formulates regional sustainable development of bamboo resources and ecological environment construction strategy by finding out the topographic Variation characteristics of bamboo resources.Fujian Shunchang is taken as an example in this paper,combined with remote sensing data and overlay analysis of terrain factor,the distribution of bamboo and dynamic change of terrain were studied by GIS and remote sensing.The results show that bamboo is mainly located at the areas where the sea level altitude is less than 1 200 m,the largest area of bamboo is concentrated in the area where the sea level altitude is from 400 m to 800 m,the bamboo area enlarged with the increasing of altitude when the sea level altitude is less than 800m,and the bamboo forest area is gradually reduced with the elevation of altitude when the sea level altitude is over 800 m; The bamboo is mainly located at the areas of which the slope is 6~25°,and mainly located in the semi-sunny and semi shaded regions.The bamboo area is rapidly growing from 1988 to 2007 in Shunchang,and the increasing area is 1 6811.48 hm2.The result reflects the dynamic area of bamboo in Shunchang which is affected by the natural and man-made driving forces,and the result provides a scientific basis for planning and ecological construction.
For the healthy and safety of the capital eco-environment,NDRC is strengthening the cooperation with neighboring regions by developing a lot of eco-environment governs to construct Beijings natural barriers and guarantee the “Green Beijing” construction.The way of inter-regional cooperation project brings great difficulty to the monitoring and management.As a new technical method,the remote sensing (RS) technology is helpful to master and evaluate the situation of engineerings implementation,and ensure the effective implementation of the environmental compensation mechanism.In this paper,firstly,business procedure re-arrangement and demand analysis are taken for the govern of eco-environment in Beijing and surrounding area invested by NDRC,such as control engineering of Beijing-Tianjin sandstorm source,“Rice to dry land crops transformation” engineering,etc.Secondly,a monitoring index system is established by using multi-source and multi-temporal RS images,and the character information of the before and after the project implementation is obtained by the combined methods of automatic classification by computer,interactive interpretation and field investigations,etc.At last,the results are analyzed to assess the practical effects of diverse control and management engineering.The research results show that:①The method can accurately locate the engineering management area,and prevent repeated and illegal engineering applications;②It is able to supervise the engineering implementation and evaluate the effect of the engineering;③It will assist to determine the ecological compensation;④It will provide a foundation data for project planning.
