Vehicle detection from aerial images has been extensively studied in many research papers and it is an important component of an intelligent transportation system. In the meantime, it is still a difficult problem with many open questions due to challenges caused by various factors such as low resolution of the aerial images, features restricted to a particular type of car, noise from other objects or object shadows, and occulsion in urban environments. By investigating several benchmark methods and frameworks in the literature, this paper proposes a novel feature fusion framework which successfully implements an effective vehicle detection method based on shadow detection followed by a rotational invariant shape matching of corner features. Promising results are obtained from the experiments.