Scene Modeling For Simulation Of Hyperspectral Remote Sensing System

Hyperspectral imaging system has the character of combining the image and spectrum. Surface spectral reflectance will vary with different material compositions, contents, existence and change of states, and spectral imaging is a discipline that studying the relationship between them. Hyperspectral imaging system has been used in many applications including object detection, material identification, land cover classification, city remote sensing and. military investigation.By building simulation models of hyperspectral system, the key contributors to system performance might be found. We can better design and operate of hyperspectral systems with the help of these simulation models. The model can be used to describe system process under various situations without real operation, this can save the cost of study.The simulation of a simple two-class scene and artificial woodland scene are discussed in detail in this dissertation. Firstly, this dissertation provides the concept of hyperspectral imaging system, then surface's absorption and reflectance of solar radiance is discussed, followed with spectrum characteristics of several typical surface classes. Secondly, the simulation of the simple two-class scene is studied. The study is focused on the surface reflectance model and solar radiance & atmosphere transmission coefficient model. Finally, the simulation of artificial woodland scene is studied based on the methodology of two-class scene model. Subpixel fraction model and secondary illumination model are discussed in detail, followed with the results of the simulation.Performances of different scene are analyzed. The results of experiment show that these models can represent the hyperspectral system process and the surface spectrum characters in a rational way.