Vehicle detection in aerial imagery : A small target detection benchmark

• We made a new database for vehicle detection in aerial imagery.
• We tested several state-of-the art detection algorithm on the database to provide a Baseline.
• Most of the best methods gives similar results.
• The best methods are still far from achieving good enough results for industrial applications.

This paper introduces VEDAI: Vehicle Detection in Aerial Imagery a new database of aerial images provided as a tool to benchmark automatic target recognition algorithms in unconstrained environments. The vehicles contained in the database, in addition of being small, exhibit different variabilities such as multiple orientations, lighting/shadowing changes, specularities or occlusions. Furthermore, each image is available in several spectral bands and resolutions. A precise experimental protocol is also given, ensuring that the experimental results obtained by different people can be properly reproduce and compared. Finally, the paper also gives the performance of baseline algorithms on this dataset, for different settings of these algorithms, to illustrate the difficulties of the task and provide baseline comparisons.