Abstract
A classification method which takes into account not only spectral but also spatial features for LANDSAT-4 and 5 Thematic Mapper (TM) data is proposed. Due to the increased spatial resolution of TM (30m compared with 80m for MSS), the number of ground cover spectral classes which are included in the Instantaneous Field of View (IFOV), decreases comparatively. This implies that spatial-spectral varibility for TM data increases in comparison to MSS. Therefore, treatment of the spatial-spectral variability existing within a region is more important. Standard deviations in small cells, such as 2x 2, 3x 3 and 4 x 4 pixels, were used as measures to represent the spatial-spectral variabilities. This information can be used together with conventional spectral features in an unified way, for the traditional classifier such as the pixel-wise Maximum Likelihood Decision Rule (MLDR). I focused my attention on the classification of new .clear cuts and alpine meadows which were very close in spectral space characteristics and difficult to distinguish them by conventional methods. There was a substantial improvement in overall classification accuracy for TM forestry data. The probability of correct classification;PCC for the new clearcuts and the alpine meadows classes rose by 7% to 97% correct. The confusion between alpine meadows and new clearcuts was reduced from 9% to 3%.
