ABSTRACT
Chapter 7 is an overview of original optical remote sensing studies conducted by the author over the years. Among regular aerial optical observations of a wind driven sea surface, wave breaking, and foam and whitecap coverage, a novel experimental material obtained from space is presented. The focus is on analysis of multispectral satellite optical imagery of high spatial resolution (~1-4 m). The research is based on a concept of the so-called optical spectral portrait. It is a spatial mosaic of specific spectral features generated from optical images of high resolution. The portrait is a product of computer vision but not a human vision. Spectral portrait allows us to select and specify areas of the interest in the images that are associated with certain ocean phenomena/events. By this means, spectral portraits perform the function of a robust classifier in computer vision of weakly emergent hydrodynamic phenomena (e.g., turbulent wake). In this context, the chapter demonstrates space-based optical capabilities to distinguish critical hydrodynamic situations (called “background” and “complex”) using a concept of spectral portrait. The published original material may provide scientific and technological breakthroughs in the field of ocean remote sensing and advanced applications. The concept and methodology reported in this chapter can be adapted on a number of other types of advanced sensors (e.g., sophisticated radar or imaging lidar) which potentially enable to provide hydrodynamic detection. Our preliminary results show that satellite quality optics with ~1 m resolution is efficient and productive tool for hydrodynamic detection. Further efforts and developments are required to elaborate operational remote sensing technology.
