Browsing by Subject "Land cover classification"

Sort by: Order: Results:

Now showing items 1-2 of 2
  • Hurskainen, Pekka; Adhikari, Hari; Siljander, Mika; Pellikka, Petri; Hemp, Andreas (2019)
    Classifying land use/land cover (LULC) with sufficient accuracy in heterogeneous landscapes is challenging using only satellite imagery. To improve classification accuracy inclusion of features from auxiliary geospatial datasets in classification models is applied since 1980s. However, the method is mostly limited to pixel-based classifications, and the coverage, accuracy and resolution of free and open-access auxiliary datasets have been poor until recent years. We evaluated how recent global coverage open-access geospatial datasets improve object-based LULC classification accuracy compared to using only spectral and texture features from satellite images. We applied feature sets topography, population, soil, canopy cover, distance to watercourses and spectral-temporal metrics from Landsat-8 time series on the southern foothills and savanna of Mt. Kilimanjaro, Tanzania, where the landscape is characterized by heterogeneous and fragmented mosaic of disturbed savanna vegetation, croplands, and settlements. The classification was based on image objects (groups of spectrally similar pixels) derived from segmentation of four Formosat-2 scenes with 8m spatial resolution using 1370 ground reference points for training, validation, and for defining 17 LULC classes. We built six Random Forest classification models with different sets of object features in each. The baseline model having only spectral and texture features was compared with five other models supplemented with auxiliary features. Inclusion of auxiliary features significantly improved classification overall accuracy (OA). The baseline model gave a median OA of 60.7%, but auxiliary features in other models increased median OA between 6.1 and 16.5 percentage points. The best OA was achieved with a model including all features of which elevation was the most important auxiliary feature followed by Enhanced Vegetation Index temporal range and slope degree. Applying object-based classification to geospatial information on topography, soil, settlement patterns and vegetation phenology, the discriminatory potential of challenging LULC classes can be significantly improved. We demonstrated this for the first time, and the technique shows good potential for improving LULC mapping across a multitude of fragmented landscapes worldwide.
  • Heinilä, Anna Maaria Kirsikka; Salminen, Miia; Metsämäki, Sari; Pellikka, Petri Kauko Emil; Koponen, Sampsa; Pulliainen, Jouni (2019)
    We aim a better understanding of the effect of spring-time snow melt on the remotely sensed scene reflectance by using an extensive amount of optical spectral data obtained from an airborne hyperspectral campaign in Northern Finland. We investigate the behaviour of thin snow reflectance for different land cover types, such as open areas, boreal forests and treeless fells. Our results not only confirm the generally known fact that the reflectance of a melting thin snow layer is considerably lower than that of a thick snow layer, but we also present analyses of the reflectance variation over different land covers and in boreal forests as a function of canopy coverage. According to common knowledge, the highly variating reflectance spectra of partially transparent, most likely also contaminated thin snow pack weakens the performance of snow detection algorithms, in particular in the mapping of Fractional Snow Cover (FSC) during the end of the melting period. The obtained results directly support further development of the SCAmod algorithm for FSC retrieval, and can be likewise applied to develop other algorithms for optical satellite data (e.g. spectral unmixing methods), and to perform accuracy assessments for snow detection algorithms. A useful part of this work is the investigation of the competence of Normalized Difference Snow Index (NDSI) in snow detection in late spring, since it is widely used in snow mapping. We conclude, based on the spectral data analysis, that the NDSI-based snow mapping is more accurate in open areas than in forests. However, at the very end of the snow melting period the behavior of the NDSI becomes more unstable and unpredictable in non-forests with shallow snow, increasing the inaccuracy also in non-forested areas. For instance in peatbogs covered by melting snow layer (snow depth <30 cm) the mean NDSI-0.6 was observed, having coefficient of variation as high as 70%, whereas for deeper snow packs the mean NDSI shows positive values.