Title of article :
Preclassification as an Aid to the Improvement of Thematic and Spatial Accuracy in Land Cover Maps Derived from Satellite Imagery
Author/Authors :
Watson، نويسنده , , Niall and Wilcock، نويسنده , , David، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2001
Abstract :
Ecological survey maps are used (1) to calibrate and validate land cover classifications from Landsat Thematic Mapper imagery and (2) as contextual information to improve spatial accuracies of the classified imagery. An objective function measures thematic accuracy, and spatial accuracies are calculated by pixel counts. Preclassification of the imagery into land classes—areas in which the combined influences of altitude, slope, aspect, geology, climate, soil, and biotic factors are believed to operate distinctively on land cover—improves classification accuracies. The usefulness of land class as a preclassifier of satellite imagery and of ecological surveys as a source of contextual information are discussed.
Journal title :
Remote Sensing of Environment
Journal title :
Remote Sensing of Environment