Optical remote sensing relies on exploiting multispectral and hyper spectral imagery possessing high spatial and spectral resolutions respectively. This book presents current state-of-the-art algorithms that address the following key challenges encountered in representation and analysis of such optical remotely sensed data.

Optical remote sensing involves acquisition and analysis of optical data – electromagnetic radiation captured by the sensing modality after reflecting off an area of interest on ground.  Optical image acquisition modalities have come a long way – from gray-scale photogrammetric images to hyperspectral images. The advances in imaging hardware over recent decades have enabled availability of high spatial, spectral and temporal resolution imagery to the remote sensing analyst. These advances have created unique challenges for researchers in the remote sensing community working on algorithms for representation, exploitation and analysis of such data. Early optical remote sensing systems relied on multispectral sensors, which are characterized by a small number of wide spectral bands. Although multispectral sensors are still employed by analysts, in recent years, the remote sensing community has seen a steady shift to hyperspectral sensors, which are characterized by hundreds of fine resolution co-registered spectral bands, as the dominant optical sensing technology. Such data has the potential to reveal the underlying phenomenology as described by spectral characteristics accurately. This “extension” from multispectral to hyperspectral imaging does not imply that the signal processing and exploitation techniques can be simply scaled up to accommodate the extra dimensions in the data. This book presents state-of-the-art signal processing and exploitation algorithms that address three key challenges within the context of modern optical remote sensing: (1) Representation and visualization of high dimensional data for efficient and reliable transmission, storage and interpretation; (2) Statistical pattern classification for robust land-cover-classification, target recognition and pixel unmixing; (3) Fusion of multi-sensor data to effectively exploit multiple sources of information for analysis.

From the reviews:“This book is composed of contributions that address many of the contemporary problems faced when processing hyperspectral image data. As expected, there are chapters focussed on thematic mapping and classification, spectral un-mixing, morphology and compression … . important to anyone interested in the state of kernel based methods in image analysis. … It is very well written, makes very good use of examples and will be an important reference work for those working on un-mixing problems.” (John Richards, IEEE Geoscience and Remote Sensing Society Newsletter, June, 2011)“This excellent reference focuses on advances in signal processing and exploitation techniques for optical remote sensing with a collection of state-of-the art algorithms for hyperspectral and multispectral imaging technologies. It is intended for advanced users, particularly graduate students and image scientists specializing in the field of optical remote sensing. … This cutting-edge publication includes a collection of images and graphics, processing technologies and parallel implementations with up-to-date references at the end of each chapter.” (Axel Mainzer Koenig, Optics & Photonics News, July, 2012)