SUPERVISED CLASSIFICATION OF MULTISPECTRA L REMOTE-SENSING IMAGE USING B-P NEURAL-NETWORK

A method of supervised classification of multispectral remote sensing image was suggested using B-P neural network. An example of supervised classification of 4 spectral remote sensing images for 4 crops was de scribed by this method. The results show that even if the feature extr actions are fewer than the traditional classification, the classificat ion of multispectral remote sensing image tallys with the cases.