Classifying remote sensing image using Gene Expression Programming algorithms

Gene Expression programming (GEP) is a new algorithm of evolutionary computation, which possesses much more powerful abilities of parallel computation and global search to resolve complex problems. This paper constructs the GEP classification model and carries out experiments with Landsat Thematic Mapper(TM) image of Dongguan city in 1997. The result shows that the GEP classification model has good classification accuracy and demonstrates effectiveness in remote sensing image classification.