One of the first texts to focus on investigating, designing and implementing algorithms and computer programs as an introduction to the rapidly evolving field of hyperspectral image and signal processing
Covering a range of applications, the authors provide a tutorial on hyperspectral image analysis, focusing on the mathematical, physical, and algorithmic models necessary to devise programs that can extract the useful information that is present in measured hyperspectral data. The amount of data produced by a hyperspectral imaging device can be enormous so care and advanced processing steps must be taken to efficiently and effectively extract information. The reader will learn about these processing steps. The authors take the readers through the topic step-by-step; from the physics foundations of the acquisition process, to the particular algorithms and families of processing tools for classification, feature selection/extraction, visualization, unmixing and classification. Homework problems are provided whereby some problems are mathematical in nature whereas others involve writing brief computer programs.
- Describes the science and hardware technology underlying hyperspectral image analysis.
- Focuses on the mathematical and algorithmic concepts for processing hyperspectral data.
- Teaches readers the conceptual basis of how the hundreds of bands in spectral pixels can be used to gather information about the materials and objects that are present in the field of view (or scene) of a hyperspectral camera.
- Outlines how to write programs that can find things that are smaller than a single pixel, and in turn details how to write programs that can describe and classify components of a scene.
- Shows how programs can use spatial information together with spectral information to produce more accurate automated analyses of images.
- Illustrates methods with a number of examples from across several applications areas, such as estimating the extent of an oil spill, detecting toxic gases around industrial plants or for homeland security, imaging human tissue to aid medical diagnosis.
- Includes companion website hosted by the authors offering publicly available hyperspectral images and sample programs for processing, as well as Matlab code.