This work focuses on neural network classifiers on sub-regions of the image. Texture information obtained with a wavelet transform is integrated in such a classifier for purposes of improving its performance. Using a wavelet transform, statistical texture information in Landsat/TM imagery is accounted. Statistical texture is extracted with a continuous edge-texture composite wavelet transform. Finally, the network is trained with both the texture information and the additional pixel labels provided by the ground truth data.
|Number of pages||13|
|Journal||Proceedings of SPIE - The International Society for Optical Engineering|
|State||Published - 1999|