Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: classification accuracy assessment

Sortuj według:

Ogranicz wyniki do:

Mapy podstawowych form pokrycia i użytkowania terenu zlewni Raby powyżej Zbiornika Dobczyckiego - porównanie dokładności klasyfikacji pikselowej i obiektowej obrazów LANDSAT TM

Badurska M., Drzewiecki W., Tokarczyk P.

Teledetekcja Środowiska

2009

T. 42

15--21

The research presented in the paper has been aimed at mapping the basic types of land-use in the upper Raba watershed (south Poland). The maps have been prepared for a study of the influence of land-use changes within the watershed on the sediment yields introduced into the reservoir. Because the erosion models used for sediment yields prediction need only to identify the main land-use / land cover classes (arable land, meadows and pastures, forests, waters, developed areas), the maps have been based on classification of middle-resolution satellite images (Landsat TM). In the research the results of traditional pixel-based classification were compared to the ones obtained in the object based approach. Six different Landsat TM images were classified. The methodology of both classification approaches have been described in the paper. The accuracy assessment of the classification results was based on their comparison with the land use types defined by the photo interpretation of colour composite images. The assessment was done by two operators. Each of them used different set of two hundred and fifty randomly generated sample points. In most cases the pixel-based approach resulted in higher overall accuracy. However, if overall accuracy confidence intervals are taken into consideration, none of the methods can be definitely recognised as a better one.

Membership function - ARTMAP neural networks

Sinčák P., Hric M., Vaščák J.

TASK Quarterly : scientific bulletin of Academic Computer Centre in Gdansk

2003

Vol. 7, No 1

43-52

The project deals with the application of computational intelligence (CI) tools for multispectral image classification. Pattern Recognition scheme is a global approach where the classification part is playing an important role to achieve the highest classification accuracy. Multispectral images are data mainly used in remote sensing and this kind of classification is very difficult to assess the accuracy of classification results. There is a feedback problem in adjusting the parts of pattern recognition scheme. Precise classification accuracy assessment is almost impossible to obtain, being an extremely laborious procedure. The paper presents simple neural networks for multispectral image classification, ARTMAP-like neural networks as more sophisticated tools for classification, and a modular approach to achieve the highest classification accuracy of multispectral images. There is a strong link to advances in computer technology, which gives much better conditions for modelling more sophisticated classifiers for multispectral images.