Hyperspectral remote sensing images incorporate properties of geometry, spectra and radiance. These features make facilitate detection and recognition of targets, and largely promote the development of target detection using hyperspectral imagery. Hyperspectral remote sensing provides detailed spectral data for target detection and recognition, the data provides our subject opportunity. However, due to detectors’ capability limitations, hyperspectral images always do not have high spatial response. Therefore, interested targets are exposed with low probability commonly. Traditional target detection methods based on spatial features are invalid for hyperspectral images. How to enhance the strong points and how to use this information is an important issue for target detection in hyperspectral imagery. Thereby it is quite necessary to make comprehensive researches on spectra enhancement and feature extraction.
Based on integrated study of the actuality of target detection using hyperspectral imagery, this dissertation analyzes four key problems in the subject, namley spectral uncertainty, dimension reduction and image denoising, mixed pixel and endmember extraction, low probability exposed target detection. The research is focused on noise estimate, data dimension reduction and feature extraction, along with that, analyzes influences of spectral uncertainty, remote sensing system and data pretreatment on target detection are also discussed. At last, more robust detection algorithms for low low probability exposed target are developed. The main aspects of this dissertation are as follows:
1. The general noise model in imaging spectrometer is discussed. To reduce the pressure of image content to traditional local means and local standard deviations method, three improved methods are developed, which are based on edges eliminate, Gaussian curve extraction and residual adjustment. At the same time, a new noise estimate method based on homogeneous region division is developed. This method is more reliable and adaptable, and works well for hyperspectral images with diverse land cover types.
2. This dissertation study widely used dimension reduction methods in hyperspectral imagery. Some key questions in maximum noise fraction are also analyzed. Based on analysis of characters of these dimension reduction methods, such as sensitivity to magnitude change and noise, information loss and data structure change, a dimension reduction method selection strategy for target detection is developed.
3. Spectral uncertainty due to human activities is studied. Results show that man-made camouflage change the spectra of the target in the full band ranges, but spectral characteristics in short-wave infrared range can be used to distinguish different materials under paints. Paints can only affect the amplitude of spectra, and the shapes of the spectra are fairly consistent. Therefore, spectral shape should be the main measurement to detect targets.
4. Hyperspectral imagery measurements and data processing are studied. Six aspects contain paints, spectral response, imaging spectrometer mode, image noise and dimension reduction are discussed. The influences of these factors to target detection are analyzed. This research provides a strategy to precision control of target detection in hyperspectral imagery.
5. The form of low probability exposed target existed in hyperspectral imagery are studied. The phenomenon of mixed pixel in hyperspectral target detection is also discussed. A low noise sensitive and partial restricted independent component analysis method is developed for decomposition of mixed pixels. Based on the idea of multiple algorithm fusion, spatial continuity, convex volume, independent component analysis and RX detector are combined together to extract endmember quickly and to detect target accurately.
With the deterioration of inland water pollution, monitoring inland water quality is becoming urgent. Monitoring water quality by remote sensing technology has the advantages of rapidness, wide coverage, low cost, and dynamic monitoring over a long period of time. However, monitoring inland water quality by remote sensing is far behind ocean color remote sensing in both development of remote sensors and monitoring approaches. The multi-spectral remote sensing data, which are often used to monitor inland water quality, can not catch the complicated and changeful spectral characteristics of inland waters accurately. Therefore, the accuracy of water quality monitoring from multi-spectral remote sensing data is quite limited. The development of hyperspectral remote sensing technique has provided much opportunity in monitoring inland water quality. Meanwhile, it has also brought challenge to traditional approaches of monitoring inland water quality. Empirical and semi-empirical approaches, which are often used in monitoring inland water quality, have low robustness, and are hard to be applied to different seasons and areas. In contrast, analytical approaches are based on bio-optical model, and have the advantages of definite physical meanings, higher robustness, and wider applicability. Therefore, it is of great significance to carry out the study on retrieval of inland water quality parameters from hyperspectral remote sensing data by analytical approach.
Taihu Lake is selected as study area in this dissertation. Based on the experiment data acquired in Taihu Lake in four seasons, five aspects of researches are accomplished: 1) measure inherent optical properties and analyze their temporal and spatial distributing rules; 2) measure apparent optical properties and analyze their spectral characteristics; 3) build bio-optical model and retrieve inherent optical properties; 4) set up and validate analytical approaches to retrieve water qualtiy parameters; 5) retrieve water qualtiy parameters from hyperspectral remote sensing image.
Main contributions of this dissertation can be concluded as follows:
1. Temporal and spatial distributing rules of inherent optical properties and specific inherent optical properties are analyzed for Taihu Lake, and on this basis, a specific inherent optical properties database is build up;
2. Four kinds of spectral indices are defined to classify water grass and algal bloom;
3. A nonlinear optimized method based on bio-optical model is proposed to calculate backscattering coefficient of suspended matter;
4. Based on the theory of matrix inversion approach, nonlinear optimized approach and algebra approach are set up, which take specific inherent optical properties as input parameters;
5. The three kinds of approaches, which are matrix inversion approach, nonlinear optimized approach and algebra approach, are tested by the four times of experiment data. The results show that retrieval accuracy of chloroyphyll-a and suspended matter concentrations are fairly good. For different seasons, the approaches to get best retrieval accuaracy of chloroyphyll-a and suspended matter are commonly not the same, and the bands combinations used in the approaches are also not the same;
6. A flow chart of retrieving water quality parameters from hyperspectral remote sensing images is designed, and a approach to calculate remote sensing reflectance is set up based on atmospheric radiant transfer model 6S. One CHRIS image is employed to test the analytical approach propsed in this dissertation, and the results are quite reasonable.
All the contributions of this dissertation have provided theoretical and methodological support in monitoring inland water quality from hyperspectral remote sensing data.
Recently, the atmospheric pollution and awful air have threaten our nation, therefore it cries for the study the atmospheric properties. As a new technology, remote sensing has a lot of advantages.
The paper selected Lake Taihu as our study area. By collecting the spectrum of the water surface, measuring of atmosphere using the sunphotometer, and acquiring the image of the MODIS on the same day, the data were used to get properties of atmosphere. The key point of the study is related to the atmosphere water vapor amount and the aerosol optical depth (AOD). The research results in this paper are presented as following:
1) Propose an improved algorithm for the retrieval of aerosol optical depth using the direct sunlight data of sunphotometer. Different effects on the inversion errors of AOD are mainly analyzed and improved algorithms have been used to get high inversion precision of aerosol optical depth, respectively in the air mass, Rayleigh optical depth, and ozone absorption coefficient etc.Especially in the computation of standard Rayleigh optical depth using the integral form, not only the effects of complex refractive index but also the depolarization ratio. The computed results have been showed that this algorithm can improve the accuracy of the retrieval.
2) A new calibrated method is presented .The different techniques of calibration for the channels of the direct sunlight are compared and applied in the actual measurements. Except of the Langley method and the standard translated one, a new calibrated one is presented in the theory.
3) As the atmosphere water vapor amount is inverted using the sunphotometer and MODIS data. The good effects are shown in the instant method of the Langley and the ratio of reflectance for two bands and the water vapor amount of every image pixel are obtained.
4) A method based on measurements for determining the aerosol type is announced. By using the nearly synchronous measurements of the data from surface spectrum and the sunphotometer with the image, and by use of the radiative transfer model 6S, varying the components of the aerosol type, a LUT (look up table) is made for the radiance on the satellite. When the total relative error of the new defined parameter for relative error is getting to the least, the aerosol type will be decided.
5)The Mie code which was used to calculate the single sphere particle has been improved, extended and realized to compute the optical characteristics including polarization of the poly –disperse aerosols. And it has been used to get the polarized phase function of aerosols over Lake Taihu.
The results of this study will be provided for the supporting of the atmospheric corrections and the water color remote sensing.