Application of neural network technique to classification of remotely sensed image

The traditional approaches of classification are always unfavorable in the description of information distribution. The artificial neural network technique developing rapidly in recent years provides a new means for ameliorating. This paper describes the BP neural network approach for the classification of remotely sensed images and attempts to develop the improved method to train the selected data. A qualitative comparison demonstrates that both original images and the classified maps are visually well matched. A further quantitative analysis indicates that the accuracy is better than the result of the traditional approaches.