Study on image-segmented classification

This paper presents an image-segmented classification method. In this method, the image is firstly divided into several segments or areas according to the spectral features, then the different training schemes are used for classification of the different segments; finally, all the results can be combined automatically into a file. The experiment for this method is land use classification for an image coming from two different scenes. The results are 6 classes in which classification precision are more than 80% and 3 classes more than 90%. However, in the classification for the whole image at a time, there are only two classes in which classification precision is more than 80%. The experiment proves that the image-segmented method can improve the quality of image interpretation in accuracy.