Some novel approaches in laser remote sensing of natural waters based on recognition of spectral patterns with the help of artificial neural networks

摘要

The results of research of application of spectral patterns recognition technique in laser remote sensing of natural waters, conducted by the authors in 2001 - 2006 years, are summarized in this paper. It is shown that application of artificial neural networks allows to perform precision analysis of water Raman scattering and fluorescence bands of humic substances (with possible distortions of those by fluorescence of pollution and other organic contaminants). Based on this, it is possible to solve such problems of laser diagnostics of natural waters as remote determination of temperature, identification and determination of concentration of salts, humic substances, oil pollutions in water, monitoring of hydrological structures.