The work presents an algorithm for recognition of isolated airstrips in multiscale Google satellite images. First, strong straight lines are detected by Radon Transform and airstrip detection is accomplished by detecting longest straight lines in multiresolution images at different altitudes. Later normalized crosscorrelation is used to find the degree of similarity among multiscale airstrip patterns. Finally, support vector machines are used to recognize airstrips. The proposed technique shows promising results in classifying airstrips from other commonly appearing objects in optical images taken from satellites i.e. roads, canals, large buildings, etc. Algorithm is dually tested on multiresolution images captured using various cameras at different heights and have produced similar results.