Spatial post-processing of spectrally classified video images by a piecewise linear classifier

Publisher Website
Google Scholar
Abstract
This study focused on investigating the possibility of combining the piecewise linear classifier (PLC) with simple spatial post-processing to improve the classification accuracies of spectrally classified images. The PLC was employed to classify video images. Spatial post-processing was then applied to the classified images. This technique was found to enhance classification accuracies considerably with only a moderate increase in processing time. Since spatially processed classification results were scene-dependent, it was important to choose an appropriate size of spatial filter for a study area. However, trade-offs between classification accuracy and processing speed with increasing size of spatial filter could be determined by trial and error.