Restoration Algorithm For Tomographic Imaging Spectrometer

This dissertation presents theories of computed-tomography imaging spectrometry (CTIS), and the experimental results of CG and OSEM algorithms, which includes the following main parts:1. The imaging spectrometry and its applicationThe development, principles, classification and application of imaging spectrometry are introduced. The concept of "data cube" is analyzed and developed. Then, the methods in which "data cube" is acquired by all the different spectral imagers are presented in details. The dispersive and interfering imaging spectrometer can obtain only a slice or a curve of "data cube" in each exposure. However, the CTIS can obtain one or more projection(s) of "data cube" in each exposure.2. The principles of computed-tomography (CT) and its applicationThis part of dissertation presents the principles of CT, which includes the main theory of Radon transform, the central slice theorem by simulated experiments and the problem of limited-projected angles .At the last of this part, the result of computer-test is listed and explains the CT theory in intuitionist.Some CT algorithms, such as CG, OSEM are also presented. We test these algorithms by simulated experiments.3. The non-scanning computed-tomography imaging spectrometer and its algorithms One kind of the CT imaging spectrometer , non-scanning computed-tomographyimaging spectrometer is presented in this part of dissertation.In this part, the principle of the non-scanning computed-tomography imaging spectrometer is listed in details. The instrument of test, and the problems of limited-projected angle are analyzed.At the last of this part, two algorithms, CG and OSEM, are presented in details, which includes the analysis of the two algorithms' equations, main theories of them. This dissertation also analyzed the astringency of these two algorithms.4. The application of conjugated gradient algorithm and expectation maximization algorithmIn the fourth part of this dissertation, the application of these two algorithms inthe test is presented and the result of experiment is analyzed in details. Most of all, some problems in the application is listed and solved.At the same time, the main programs are listed in this part, and also analyzed in details. The "data cube" is showed to explain the result from intuitionistic angle.Some answers to the problems such as imputes of initialized "data cube", and the computed data, are also presented.