Based on researches on the physical mechanism of atmospheric delay, this paper has studied the atmospheric effects in repeated track InSAR measurements and put forward four ways to analyze the effects based on SAR interferograms. The research progress, especially in the correction of atmospheric effects, is reviewed comprehensively. Different approaches to mitigating the atmospheric effects as well as their problems and limitations are discussed and compared with each other. At last, some suggestions are given for future studies.

 Wave bands and the combination of RS images can reflect the spatial information of crop growth, monitor wheat growth and estimate its yield.　Including the effects of climate, soil, cultivars and cultivation measures on physiological processes such as phonological development, photosynthesis, organ building, dry matter partition and quality formation, the wheat growth model is a dynamic mathematical model with predictability and mechanism. On the basis of a brief introduction to the wheat growth model and RS monitoring, the research advances in wheat growth monitoring combining the wheat growth model and RS monitoring are summarized, and some ideas for further researches are discussed in this paper.

The Bowtie effect is a kind of geometry distortions of the EOS/MODIS L1B data. Although several methods have been proposed to eliminate the Bowtie effect, there exist many limitations in their efficiency and practical application. This paper makes a thorough research on the factors which result in the Bowtie effect. With a comprehensive comparison and evaluation of the existing algorithms for eliminating the Bowtie effect all over the world, the paper offers a new algorithm which is succinct and effective in eliminating the Bowtie effect without utilizing the ephemeris data. The new algorithm is derived from the regularity shown by the experiments made in China, the USA and Australia (mainly with MODIS data of 500 m resolution). The examination results demonstrate that, without ephemeris data, the new algorithm is capable of removing the Bowtie effect of MODIS L1B data effectively and rapidly.

 Using Landsat TM imagery, the authors designed an atmospheric correction method for nonhomogeneous atmosphere in this paper. The method is based on the traditional dark object subtraction (DOS) technique. In combination with the atmospheric radiative transfer model, the authors selected quite a few dark objects from Landsat TM imagery and calculated the atmospheric correction factor corresponding to each dark object. Then the atmospheric correction for entire image was carried out by making use of all dark objects synthetically. The main characteristic of this method is that it does not depend on any extraneous information. As the atmospheric nonhomogeneity is considered in this method, the precision of atmospheric correction is relatively high.

The false topographic phenomenon is a common phenomenon existing in the remote-sensing images obtained by sun-synchronous satellites, which brings great trouble to image users. In order to remove the false topographic phenomenon of the remote-sensing images，this paper put forward a method based on DEM data and IHS transformation and made a case study of the image of Guanyuan City in Sichuan Province. By low-pass filtering of the intensity image obtained by IHS transformation, the reflectivity information (IR) was extracted from the intensity image. By adding IR to shade relief image (SR), which was produced by DEM, a new intensity image (Inew) was obtained. A back IHS transformation was done to acquire corrected RGB image after replacing the old intensity image by the new intensity image (Inew). The experimental results indicate that the method can remove the false topography effectively and preserve the primary color.

 Roads can be represented by straight lines or linear bands conveniently. In a road image, a straight line may contain some fissures while a linear band may contain some holes and have some attached irregular blocks. The utilization of existing approaches to extract such features has some obvious defects. In order to overcome these disadvantages, this paper puts forward a new method for line extraction. The method first extracts and thins edge points, then finds a base line segment according to the structural characteristics, and finally predicts and detects the following points iteratively by making use of the obtained part. The same principle can also be used to extract the linear bands of which a road consists. The experimental results show that this method has excellent performance.

Withthemaincityzonesastheresearcharea,thispaperstudiedthemethodforextractingwaterinfor
mationinSPOT5image.First,theimageissegmentedbymulti-segmentationmethods.Afterthat,incombina
tionwithsomespectral,shapeandtexturalcharacteristics,themulti-levelclassificationmethodcanbesetup.
Theresultshowsthatthiskindofclassificationmethodusedtoextractthewaterinformationincityzoneshasnoton
lyabundantrules-meaninginformationthatcanreducethesalt-phenomenonbutalsoobviouslyhighaccuracy.

Becauseofthehighdatadimensionalityofhyperspectraldata,itisverydifficultforconventionalmeth
odstoobtainsatisfactoryresultsinthestudyofimageclassificationformaterialsontheground.Intheprocessof
featureimagesextractionbasedoninformationfusion,anewmethodisputforward,whichcombinesimage
multifeatureinformationfusionwithvariousclassificationalgorithmstomakeobjectorientedhyperspectralimage
classification.Experimentsshowthatimagedatadimensionsarereducedandclassificationaccuracyisimproved.

Segmentation of gradient images in the traditional watershed algorithm usually has no markers, which is likely to cause excessive segmentation. This paper presents a watershed algorithm based on the label. First, the gradient images are obtained through the reconstruction of gradient by using Sobel operator and, at the same time, the perimeter, area and morphology factors of the region are computed and labeled. Then, the initial image of segmentation is acquired by using the method of distance function icon for determining the seeds and the technique of isoline tracking. Finally, the last result of segmentation is obtained by using an improved method of regional merging. The experimental result shows the effectiveness of the method.

 Based on simulations using the SAIL bidirectional canopy reflectance model coupled with the PROSPECT leaf optical properties model, the authors have obtained the vegetation LAI. The FLAASH model is used to ensure the accuracy of atmospheric correction. The LOPEX93 database and the JHU spectral library are employed to identify the input value of the model. The canopy reflectance is the key node to relate the PROSPECT+SAIL model and the remote sensing data, and hence the LAI can be retrieved from leaf biochemical properties, LAD and remote sensing data. In situ measurements of LAI are used to test the simulation results, and the error sources are analyzed.

Using a canopy radiative transfer model, PROSAIL, the authors introduced soil reflection index（SRI） to simplify model, and proposed a method for computing SRI directly from reflection. Besides, sensitivity analyses of various vegetation parameters on modeling performance under different band integration approaches were conducted. On the basis of sensitivity analyses of the model, a set of new band integration approaches with genetic algorithm was induced to calculate the estimating values of LAI for Landsat TM data . Experiments with Landsat TM data indicate that the retrieval accuracy is relatively high for vegetation with LAI less than 3, and that, with LAI more than 3, the retrieval accuracy is low. These phenomena are attributed to the fact that the canopy reflection is no longer sensitive to LAI when the vegetation is too densely developed. From this study, it is concluded that LAI retrieval with the PROSAIL model is only credible in a certain range.

An improved image fusion method based on à trous wavelet decomposition is presented in this paper. The method was applied to panchromatic and multispectral images from Landsat 7 ETM+ in the Jiaozhou Bay. The multispectral images and panchromatic image were decomposed by à trous wavelet transform. The low frequency parts of multispectral images were modulated by texture properties of the panchromatic image. The wavelet coefficients were obtained by using the fusion rules based on the area-based energy ratio and maximum absolute value. At last, the fused image was reconstructed by inverse wavelet transform. Quantitative assessment of the results shows that the fused image obtained by the method has the merits of better preservation of image definition and less loss of spectral information. The fused image with high spatial and spectral information can increase the accuracy of classification.

The Principal Component Analysis (PCA) image fusion method has been used widely in recent years. However, without considering the effects of noise on the fusion image, its application is only limited to the fusion region. The Minimum Noise Fraction (MNF) transform is a self-contained component analysis method which considers the effects of noise on the fusion image. This technique is employed in such fields as the determination of the inherent dimensionality of image data and segregation of noise in the data; nevertheless, it is not applied to image fusion nowadays. Therefore, in view of the defectiveness of the PCA image fusion method and the superiority of the MNF transformation, the authors put forward a new MNF transform Remote Sensing fusion method in which both IKONOS multi-spectral image and panchromatic image are used. Visual and quantitative comparison demonstrates that this technique is better than other fusion methods.

 The locations of waterline in remotely sensed images are mainly affected by high concentration suspended sediments and surface remnant water. A decision tree model considering the water depth was applied in this paper to detecting waterline. Furthermore, waterline was also traced from the reference digital terrain model (DTM) and the associated tidal elevation. The two approaches were both used to delineate the waterline in the Yangtze Estuary, and the experimental results indicate that they are fairly effective in waterline extraction.

 Based on spatial continuous datasets (NDVI) and local indicators of spatial association (LISA)——local Moran’s I (LMI), this paper deals with the process of landscape pattern fragmentation in a small watershed by detecting the change of local spatial autocorrelation of landscape features. As a spatial explicit indicator evaluating the landscape heterogeneity, LMI can detect the hot spots during the landscape fragmentation process. The relationship between the landscape pattern change and the land use / land cover change in the past 30 years is also studied, and the results show that the hot spots of landscape fragmentation are located in downstream areas where land use changes are very dramatic. The researches performed by the authors show that the land use change in the small watershed is the main driving force resulting in such changes of landscape pattern as landscape fragmentation. Studies show that the local indicators of spatial autocorrelation are helpful to the traditional analysis of landscape pattern.

 To investigate the REE mine site exploitation in some places, the authors made remote sensing change detection by using Landsat ETM+ (1999) and SPOT (2005) dataset. Based on the results of unsupervised classification, this paper used the “post-classification comparison” change detection technique to analyze the change details, and formulated an accurate “From-to” change matrix. The increased parts of mine sites class were detected, which resulted from the decrease of other classes. Then the distribution of the change parts was delineated on the map. Before change detection, a radiometric normalization method, i.e.,“Scattergram-Controlled Regression”, was introduced to correct the spectral difference of the two temporal datasets. The accuracy assessment was also used to guarantee the acceptability of the classification results. All of these procedures could enhance the accuracy of the final results and lead to satisfactory result.

 Based on the field investigation into the Dexing copper mine, the authors drew the spectral curves for the mine with advanced hyper-spectral field testing instruments, and further realized the pollution and the diagnostic remote sensing spectral features of the water. Combined with the characteristics of ASTER and based on the theory of tricolor, this paper carried out false color composition to distinguish water pollution types and degrees effectively. The result can be regarded as a valuable reference for other similar applications of the satellite data.

EO-1 Hyperion data is used to estimate nitrogen concentration of Guanxi pomelo in this study. Based on analyzing the data characteristics, the Hyperion preprocessing includes such means as atmospheric correction and geometric correction. Using the linear stepwise regression method, this paper established the correlation between the reflectance spectra / derivative spectra and the concentration of nitrogen on the basis of the sampling data in the field. The results show that reflectance spectra derivative of logarithm is fairly good in estimating nitrogen concentration. The absorption features around 1 003 nm, 1 245 nm, 1 336 nm and 2 264 nm are selected. The values of nitrogen concentration through estimation are quite consistent with those of field measurements. The authors have thus concluded that it is feasible and time-saving to estimate nitrogen concentration of pomelo by using hyperspectral remote sensing images.

 The chlorophyll and the amount of vegetation derived from the Vegetation Index can be used to estimate the effect of ecological restoration. The traditional remote sensing data is not precise enough due to the influence of non-vegetation information. In this paper, Vegetation Index is calculated from hyperspectral remote sensing data and is used to estimate the effect in the tailing of Gejiu tin mine. Compared with the traditional method, the hyperspectra data can improve the precision by removing the non-vegetation information.

 This paper has discussed the problem of the fusion of the multi-spectral and panchromatic remote sensing images, studied the fusion method based on the IHS transform and Wavelet Packet Transform, and put forward the fusion method based on the IHS transform and Best Tree Transform. First, the multi-spectral remote sensing image is transformed with the IHS transform. Second, the panchromatic image and the I component of the multi-spectral image are transformed with the Wavelet Packet and the Best Tree Transform and then merged. At last, the merged image is obtained by inverse IHS Transform. The fusion image is fairly clear in vision and at the same time makes the entropy greatest.

 With the Daxinganling area as the experimental site, the authors used the PCA (Principal Components Analysis) to get the first weight and the results of the NDVI (Normalized Difference Vegetation Index), and employed a certain band in the LandSat TM1 to 7 to realize the combination of the wave band, which could enhance the contrast between the forest information and the background information. The study also used the method of maximum likelihood to realize the supervised classification of the images, whose accuracy could exceed 92%. This paper presents two experimental investigation images with different quantities of cloud and different extents of forest coverage. This investigation shows that this method can improve the automation and precision in the extraction of the forest coverage information.

This paper has discussed the application of the remote sensing techniques to the investigation of the forest resources in West Tianshan Mountain on the basis of CBERS-2 data. Based on the construction of the monitoring index system, the ground investigation and remote sensing analysis, and using the method of system analysis, the authors extracted the forest resource information. It is shown that the greatest proportions of the forest resources in the West Tianshan Mountain are forest lands and bush forests. It is concluded that the forest resource monitoring of the forest zone in the Tianshan Mountain by using CBERS-2 data can yield satisfactory results.

The PCI Geomatica software of Canada widely used in China allows users to define a coordinate. As much of the image processing work is based on Xi'an 80 coordinate, it is very important for us to define a coordinate needed in daily work. This paper has successfully defined the Xi'an 80 coordinate projection parameter which can ensure the effectiveness in practical work.