Using Genetic Algorithms to Analyze Multi-Spectral Imagery
Invented in the 60's to model natural selection, genetic algorithms have more recently become popular in the artificial intelligence community as an optimization technique. A multi-spectral imaging sensor simultaneously scans the target at multiple wavelengths. (The jargon-challenged sometimes refer to the resulting data as a color picture.) We will discuss how one can use genetic algorithms to convert multi-spectral images from the Landsat satellite into land-cover maps.